1 UNITED STATES DISTRICT COURT FOR THE SOUTHERN DISTRICT OF CALIFORNIA 2 VOL. I 3 PGS. 1-155 EXHIBITS 1 4 ANDREW SHALABY and ) 5 SONIA DUNN-RUZ, ) ) 6 Plaintiffs, ) ) 7 vs ) ) 8 IRWIN INDUSTRIAL TOOL COMPANY, ) INCORPORATED; THE HOME DEPOT, ) 9 INCORPORATED; and DOES 2 Through ) 100, Inclusive, ) 10 ) Defendants. ) 11 AND RELATED THIRD PARTY CLAIMS ) 12 13 DEPOSITION OF THOMAS W. EAGAR, Sc.D., an expert witness called by the Plaintiffs taken pursuant 14 to the Federal Rules of Civil Procedure, before Lori J. Atkinson, a Court Reporter and Notary 15 Public in and for the Commonwealth of Massachusetts at the offices of Synergy 16 Executive Suites, Ten Post Office Square, Boston, MA, on the 10th of September 2008, 17 commencing at 9:30 a.m. 18 19 20 21 ATKINSON-BAKER, INC. COURT REPORTERS 22 (800) 288-3376 www.depo.com 23 FILE NO.: A206C4A 24 1 1 APPEARANCES: 2 Mark D. Epstein Esq. ALBORG, VEILUVA & EPSTEIN, LLP 3 200 Pringle Avenue Suite 410 4 Walnut Creek, CA 94596 925.939.9880 5 Counsel on behalf of the Plaintiffs 6 7 Lowell Carruth, Esq. McCORMICK BARSTOW LLP 8 5 River Park Place East Fresno, CA 93720 9 559.433.1300 Counsel on behalf of the 10 Third Party Defendant, Western Industries 11 Beth Schneider Naylor, Esq. 12 FROST BROWN TODD 2200 PNC Center 13 201 East Fifth Street Cincinnati, Ohio 45202-4182 14 513.651.6800 Counsel on behalf of the Defendants 15 Newell Rubbermaid, Incorporated and Rubbermaid, Incorporated 16 17 Richard A. Ergo, Esq. BOWLES & VERNA LLP 18 California Plaza 2121 N. California Blvd. 19 Suite 875 Walnut Creek, California 94596 20 925.935.3300 Counsel on behalf of the 21 Third Party Defendant, Worthington Industries 22 Also Present: Michael Ridley 23 24 2 1 INDEX 2 Witness Direct 3 THOMAS W. EAGAR, Sc.D., 4 (By Mr. Epstein:) 4 5 6 7 8 9 E X H I B I T S 10 No. Description Page 11 1 Report 6-23-08 5 12 2 Letter 7-28-08 12 13 3 DOT-SP 6686 12 14 15 16 ***Exhibits were retained by the Court reporter 17 and attached to the transcript.*** 18 19 20 21 22 23 24 3 1 P R O C E E D I N G S 2 THOMAS W. EAGAR, Sc.D, 3 having been satisfactorily identified and first 4 duly sworn by the Notary Public, was examined 5 and testified as follows: 6 DIRECT EXAMINATION 7 BY MR. EPSTEIN: 8 Q. Will you state your full name for the record? 9 A. Thomas W. Eagar. 10 Q. Good morning, Dr. Eagar. I'm Mark Epstein. I 11 represent the Plaintiffs Andrew Shalaby and Son 12 Dunn-Ruiz in this lawsuit. I have had an 13 opportunity to review your resume' or your 14 professional biography that was attached to your 15 report. 16 Is it fair to say you have had your 17 deposition taken many times before? 18 A. Yes. 19 Q. Approximately how long have you been doing work 20 as an expert witness? 21 A. First time was probably the late '70s, so 30 22 years roughly. 23 Q. Are you currently a full-time professor at MIT? 24 A. Yes, I am. 4 1 Q. Can you tell me approximately what percentage of 2 the time that you spend in professional 3 endeavor, you split between being a university 4 professor and an expert witness? 5 A. Well, I can't -- I've never broken out expert 6 witness from what MIT calls outside professional 7 activities. But half my time is spent at the 8 university and half my time is spent on outside 9 professional activities which includes 10 industrial consulting, expert witness, serving 11 on national committees, professional society 12 committees; and things like that. 13 Q. I'm going to attach as Exhibit 1 a copy of your 14 report dated June 23, 2008? 15 MR. EPSTEIN: Anybody need copies? 16 (Document marked Exhibit No. 1 for 17 identification.) 18 BY MR. EPSTEIN (CONT'D): 19 Q. Doctor, I see you are flipping through it. The 20 copy that you have. Will you make sure that is 21 a complete copy of your report? 22 A. Yes, it is. 23 Q. I notice that your report is on MIT letterhead 24 or you use it on the top of your report. Do you 5 1 have a consulting business separate and apart 2 from your professorship at MIT? 3 A. This is not MIT letterhead. It doesn't have the 4 MIT seal on it. This is my personal letterhead. 5 The largest font is my name. 6 Q. You use address as your MIT work address? 7 A. That is where I do most of my work, yes. To 8 answer the other question yes, I do have a 9 separate sole proprietorship, which is schedule 10 C on my income tax. 11 Q. All right. Do you maintain a separates office 12 other than what is shown on the report? 13 A. No. 14 Q. In addition or -- strike that. 15 Aside from this case, how many other 16 litigation matters -- in how many other 17 litigation matters have you been retained by 18 Worthington Industries? 19 A. There are two others currently. 20 Q. Do you know the names of those? 21 A. One is Barrett and the other is Glenn. 22 Q. In those -- in addition to those two current 23 cases, how many other litigation matters have 24 you been retained by Worthington Industries in 6 1 the past? 2 A. One other. 3 Q. Do you know the name of that one? 4 A. Sandoval, S-A-N-D-O-V- A-L OR O-L. 5 Q. Is it your understanding that the Sandoval 6 matter settled? 7 A. It is my understanding that Worthington was 8 released from that. 9 Q. Okay. Other than the three matters that you 10 just listed, have you been retained by 11 Worthington Industries for any other purpose? 12 A. No. 13 Q. Have you ever been retained in any capacity by 14 Irwin Industrial Tool company before? 15 A. No. 16 Q. How about Western Industries Incorporated? 17 A. Yes. 18 Q. Can you tell me approximately how many times you 19 have been retained by Western Industries? 20 A. Well, I'm going back eight or ten years ago, 21 probably three or four times. 22 Q. Were those litigations matters or nonlitigation? 23 A. The original two or three were litigation. 24 There might have been one or two where I was 7 1 basically just helping out another expert. 2 Q. Can you tell me approximately how long ago the 3 most recent retention was by Western Industries, 4 estimate? 5 A. Four years, five years, I would think. 6 Q. You haven't been retained by Western Industries 7 in the last two years? 8 A. About two years ago, an attorney called up and 9 said they wanted to retain me. They never sent 10 me anything or anything else. Nothing has 11 happened. I probably couldn't even track it 12 down. 13 Q. The matters that you were retained -- on which 14 you were retained by Western Industries, did 15 those involve MAPP gas cylinders? 16 A. No. 17 Q. Do you recall what matters -- what they 18 entailed? 19 A. They were propane cylinders. 20 Q. Were any of them one pound propane cylinders? 21 A. Yes. They were all NRT one pound cylinders. 22 Q. In addition to the -- strike that. 23 The three litigation matters that you 24 mentioned in which you have been retained by 8 1 Worthington Industries in recent years, were 2 those -- was Attorney Rich Ergo involved in all 3 three of those? 4 A. He was involved in all of those, yes. 5 Q. Have you been retained by Mr. Ergo or another 6 attorney in his firm for any non-Worthington 7 related matters? 8 A. No. 9 Q. Can you give me an estimated breakdown of the 10 approximate half of your time that you spend on 11 outside consulting, can you give me a breakdown 12 of that time with regard to litigation versus 13 non-litigation matters? 14 A. I would say that about slightly less than half, 15 let's say 40 percent, which would be 20 percent 16 of my total time, would will be litigation 17 matters. The other 30 percent or 60 percent - 18 30 percent of my total time, 60 percent of why 19 outside professional is divided between 20 industrial consulting and professional society 21 activities. 22 Q. Of your litigation time, can you tell me what 23 percentage of that time is spent -- strike that. 24 Of your litigation time, what 9 1 percentage of that have you been retained by the 2 defense side? 3 A. It's about 50/50 any time I have gone through it 4 for more than a year averaging. If you take a 5 small little time frame, which you will find one 6 way or the other. But any time I average for 7 over more than a year, it always been 50/50. 8 Q. When were you first contacted about this case? 9 A. I believe it was last fall sometime; might have 10 been last summer. About a year ago; 10 months, 11 12 months ago. 12 Q. Were you contacted by Mr. Ergo? 13 A. Yes, I was. 14 Q. What did he tell you when he contacted you? 15 A. He told me he had several MAPP gas cases. He 16 asked me if I had ever been involved in such 17 things before. And I told him that I had worked 18 not on MAPP gas but worked with Western 19 Industries. And that was basically the extent 20 of the first conversation as I remember it. 21 Q. Were you retained for several matters at one 22 time? 23 A. I wasn't retained at that time. He wanted to 24 check out -- when he found out that I worked for 10 1 Western, he wanted to find out if there was any 2 conflict of his hiring me. 3 Q. At some point thereafter, you were retained in 4 connection with this case? 5 A. Yes. 6 Q. At the same time were you retained in connection 7 of any of the other three matter that you 8 listed? 9 A. When I was finally retained, which may have 10 taken a couple of months, I was essentially 11 retained on probably two or three at the same 12 time. I don't remember -- I think Shalaby was 13 one of the initial ones. I don't remember if 14 Barrett was involved. I can't remember if Glenn 15 or Sandoval were among the initial ones. 16 Q. Were you asked to prepare some opinions in 17 connection with this case? 18 A. I was -- I'm not sure I understand the question 19 exactly as you phrased it. I was asked to 20 investigate what happened in this matter. 21 Q. Were you asked -- you were asked to investigate 22 what happened and were you asked to render an 23 opinion as to what you believe happened in this 24 matter? 11 1 A. I eventually rendered an opinion and wrote a 2 report, yes. 3 Q. The opinion that you rendered is that set forth 4 in your report dated June 23? 5 A. Yes. The initial report. There is a second 6 supplement. 7 Q. That is a supplement dated July 28, 2008? 8 A. Yes, sir. 9 MR. EPSTEIN: I will attach that as an 10 exhibit in a moment. 11 BY MR. EPSTEIN (CONT'D): 12 Q. Other than the opinions that you have expressed 13 in your report and your supplemental report, 14 were you asked to render opinions on any other 15 subjects in this matter -- in connection with 16 this matter that are not set forth in your 17 reports? 18 A. I don't believe so. 19 MR. EPSTEIN: I am going to mark your 20 supplement report as an exhibit. Exhibit 2. 21 (Document marked Exhibit No. 2 and 3 22 for identification.) 23 BY MR. EPSTEIN (CONT'D): 24 Q. If you would take a moment to look at that and 12 1 tell me if that is a and correct copy of your 2 report? 3 A. It has an e-mail on the back which is not part 4 of mine. Other than that, it is all three pages 5 of my report. 6 Q. All right. By the way, Dr. Eagar, in your 7 initial report, starting -- first time I see on 8 Page 2, you make reference to NFPA 921? 9 A. Yes, sir. 10 Q. Can you tell me what is NFPA 921? 11 A. It is a guide to explosion and fire 12 investigations put out by the National Fire 13 Protection Association in Quincy, Massachusetts. 14 Q. Did you have any input into this document? 15 A. No, I did not. 16 Q. Are you a member of the NFPA? 17 A. No, I am not. 18 Q. Do you consider NFPA 921 to be a scientific 19 document? 20 A. It has some scientific things in it. It has 21 other things that are not science, per se, they 22 talk about litigation matters, and things like 23 that, unless you want to talk about judicial 24 science. They are technical things in there 13 1 that relate to science, yes. 2 Q. You make reference to what looks like the 2004 3 edition? 4 A. Yes. 5 Q. To your knowledge has the 2004 edition been 6 superceded by a more recent edition? 7 A. There is a 2008. I haven't purchased it yet 8 myself. 9 Q. Is NFPA 921, would you consider to be an 10 evolving document? 11 A. Yes. 12 Q. Do you know how often it is updated? 13 A. Well, all I can say is periodically, but I have 14 seen some of the notices or correspondence. 15 Because anyone in the country that wants to can 16 write to the NFPA an propose a change. And 17 there are committees or subcommittees that 18 review these proposed changes and then will 19 issue opinions on whether they are going to be, 20 I think professionally accepted, which means 21 they will go out for a review among -- a peer 22 review among a bunch of fire investigation 23 professionals or they will be rejected by the 24 committee. Then there is a whole process, which 14 1 I'm not particularly familiar with. I have seen 2 some of these documents in other matters. 3 Q. Are you on a mailing list for NFPA? 4 A. No. Well, I get their catalog, so I can 5 purchase their periodicals, which I have a 6 number of. I often have my -- the engineer who 7 runs my lab, who is a member of NFPA often have 8 him purchase them because he gets a member 9 discount. 10 Q. Do you use NFPA 921 in teaching any of your 11 classes? 12 A. I wouldn't say I've never mentioned it. It is 13 certainly not part of the curriculum, per se. 14 Q. Does NFPA 921 speak to the issue of mechanical 15 failure analysis? 16 A. I'm not sure what you mean by mechanical failure 17 analysis. It does have Chapter 4, which I 18 referenced here, which is the scientific method, 19 which is how to conduct a failure investigation 20 if that is responsive to your question. 21 Q. Okay. Does it specify or limit the types of 22 failures that it is applicable to? 23 A. Chapter 4 is very generic to the scientific 24 method. I actually use that chapter in many 15 1 other cases have nothing to do with fire or 2 explosions just because it is one of the best 3 two or three page treatises on the scientific 4 method that I have seen. It is well recognized, 5 peer-reviewed, and I will often, just as I did 6 here, use it to point out this is the way a 7 scientist goes about a forensic analysis or a 8 forensic engineer should be going about an 9 analysis. Whether it's a fire or an explosion. 10 Chapter 4 is -- that is what I use Chapter 4 11 for. Much broader than the intent of NFPA 921 12 itself as a whole document. 13 Q. You use it for more of a -- I don't know if 14 generic is the right term -- a broader scope of 15 explaining the scientific analysis? 16 A. I have other references to scientific method. 17 But I actually find this one to be a little more 18 detail. It also goes in to the determination of 19 cause in a forensic sense. In a detail that I 20 haven't found outside of some other textbooks on 21 fire investigation. The fire investigators have 22 sort of, in the last 12 to 15 years, adopted the 23 scientific method for their forensic 24 investigations. But it is actually the 16 1 appropriate method for a scientific 2 investigation of any forensic investigation, 3 which is the type of thing that I do. 4 So I refer to chapter 4 quite often. 5 Q. So you use Chapter 4 in or -- strike that. 6 Do you consider Chapter 4 of NFPA 921 7 part of your analysis in reaching the 8 conclusions that are set forth in your initial 9 report in this action? 10 A. I'm not sure I understand the question exactly, 11 but I certainly use the scientific method in 12 going about this report. In fact, the reason it 13 is number one is because I think if you follow 14 one and two basically are an outline for the 15 following parts of my report; subsequent parts 16 of the report. 17 Q. Okay. Did you use sections one and two of NFPA 18 921 in preparing your report? 19 A. I meant sections one and two of my report on 20 page two. Sections one and two of NFPA 921; one 21 is the introduction and two I think is 22 definitions. I was talking about one and two on 23 page two of my report. 24 Q. Can you tell me the chapters of NFPA 921 that 17 1 you relied in preparing your report? 2 A. In this one is really just chapter 4, we don't 3 have any physical evidence in this matter to go 4 and look at burn patterns or things like that. 5 Q. Did you refer to or rely on section 13 -- or 6 Chapter 13 of NFPA 921? 7 A. I would have to go look at Chapter 13. Over the 8 years, NFPA 921 has changed the chapter headings 9 and numbers. I think originally Chapter 4 was 10 Chapter 2 in an earlier edition. 11 Q. If I told you that chapter 13 deals with witness 12 testimony -- 13 A. I do know that some of the other experts in this 14 case have been opining about truthfulness of 15 witness testimony, which I don't -- I haven't 16 reviewed all of Chapter 13 in NFPA 921 for this 17 matter. But I actually found it inappropriate 18 for forensic experts to be opining on the 19 truthfulness of witnesses. They can talk about 20 witnesses have said and whether that is evidence 21 that they relied on, but to opine on the 22 truthfulness of it, seems to be well beyond the 23 role of an expert in a forensic investigation. 24 Q. Where did you get the impression that other 18 1 witnesses have opined on the truthfulness of 2 witnesses in this case? 3 A. I have read the depositions of Dr. Anderson and 4 and Dr. -- I'm going to murder this -- 5 Vrendenburgh. And they were basically saying 6 that certain witnesses couldn't be true. And 7 they were discounting those -- that evidence 8 totally, which seemed to me to be violating 9 Chapter 4 of the NFPA guidelines, which you have 10 to at least consider all the evidence. You may 11 end up coming to a conclusion that might be in 12 contradiction to some of the evidence but you 13 can't just discount it, because you don't 14 believe it is truthful. 15 Q. Is it your understanding that Dr. Anderson 16 concluded in his mind that certain witness 17 testimony was truthful and other witness 18 testimony was not truthful. 19 A. I think he used the word bazaar. Certain 20 testimony was bazaar. I think to characterize 21 testimony with such pejorative terms is 22 inappropriate. Okay? He may say he didn't want 23 to rely on it or something, that's fine. But to 24 start using pejorative terms like bazaar seems 19 1 to be somewhat of an advocacy position to me. 2 Q. What about making assessments of witness 3 reliability, is that appropriate in, for 4 example, fire cause and origin investigation? 5 A. Certainly looking at evidence and considering 6 all of it and trying to determine how all of it 7 fits together in a whole requires that an 8 investigator determine which of the facts, which 9 are sometimes in conflict, seems to present the 10 most consistent reliable pattern. So therefore, 11 while you should consider everything, you don't 12 have to rely on other things. 13 Q. Okay. So then is it appropriate for the 14 investigator to make an assessment with regard 15 to what evidence in their mind might be more 16 reliable than other evidence as presented? 17 A. Well, I always end up kind of coming to an 18 overall opinion and therefore I end up -- if you 19 go backwards, you will find that certain things 20 I have deemed to be more reliable than others. 21 But that wasn't something where I didn't 22 consider -- to begin to consider it. 23 Q. But the assessment as to what may be more 24 reliable than others -- what evidence might be 20 1 more or less reliable than other evidence, that 2 is inherent in part of what you do as a forensic 3 investigator, isn't it? 4 A. You end up relying on some facts more than 5 others based on whether they are any consistent 6 with other facts. Rarely are all the facts just 7 crystal clear and undisputable. But you should 8 consider everything, and I think as NFPA 921 9 says, as you consider it, you have to do 10 something to rule things in or rules things out. 11 But to just say something is bazaar or you 12 wouldn't expect someone to do something foolish 13 in front of their children. In my experience I 14 have seen lots parents do foolish things in 15 front of their children. I have probably done 16 foolish things in front of my children. Okay. 17 Q. Did you happen to bring a copy of NFPA 921 with 18 you? 19 A. No, I did not. 20 Q. Are you familiar -- I'm going to quote a portion 21 of NFPA 921 from Section 13 -- Chapter 13 22 Section 13.1.2.1, which begins, "Generally, any 23 information solicited --" 24 (Interruptions in proceedings.) 21 1 Q. 13.1.2.1 has a sentence which reads in part, 2 "Generally any information solicited or received 3 by the investigator during a fire investigation 4 is only as reliable as the source." Are you 5 familiar with that? 6 A. That seems like almost a tautology. 7 Q. Like a what? 8 A. A tautology is something that is equivalent. 9 Substitute and equal sign or the word is. 10 Q. Okay. So then if I understood you correctly, in 11 conducting a forensic investigation or analysis 12 into the failure cause of a fire, it is 13 appropriate for an investigator to make certain 14 determinations along the way in terms what 15 evidence that is presented is reliable and what 16 evidence is not as reliable or not perhaps 17 reliable at all? 18 A. Certainly, as I said before an investigator will 19 ultimately in their opinion have chosen a 20 subsequent of facts in which some are considered 21 more reliable than others. Just because what 22 their conclusion is and the fact that you had 23 conflicting facts to begin with, if you draw a 24 conclusion then inherently some things are going 22 1 to be -- have been considered more reliable than 2 others. If you are doing a scientific 3 investigation, you should have a scientific 4 rather than an emotional basis for choosing 5 that. 6 Q. How about common sense, is an investigator 7 supposed to use common sense in conducted a fire 8 cause and origin analysis? 9 A. We are getting into tautologies. It is like 10 motherhood and apple pie. Certainly, who is 11 going to argue against common sense in an 12 investigation. I can't imagine anyone arguing 13 against common sense. 14 Q. Okay. Referring to your June 23rd report. In 15 your first section there, you make reference to 16 -- strike that. 17 Did you, in conducting your report, 18 try to acquire some sort of numerical results, 19 some form of analysis? 20 A. When I could, but without any physical evidence, 21 it is clearly more difficult. We do some 22 numbers like two and a half inch long crack or 23 split. You will see I have some numbers in 24 here. But the numbers are really sort of 23 1 generic numbers on the design because we don't 2 actually have the physical evidence. 3 Q. And referring you to section 3 of your report on 4 page 3. You make reference to CFR, section 5 178.65 is that the DOT regulation, that 39? 6 A. Yes, it is. 7 Q. Is it your opinion that if a product meets the 8 standards of -- refer to as CFR 178.65 that in 9 and of itself that the product is necessarily 10 free of any design flaws? 11 A. Well. I don't think you can make a generic 12 statement like that without looking at the 13 history. However, we do have a history here and 14 this particular specification 39 goes back 15 decades so far as that goes. There are 10 of 16 millions if not hundreds of millions of 17 cylinders that have contained propane and MAPP 18 and things over the years. So we have a 19 considerable history on the quality of this 20 design. So I'm not trying to fight with you, I 21 am just saying -- the way you made the 22 statement, I can't just take that as a general 23 statement. But if I combine that statement with 24 the history of these types of cylinders and this 24 1 specific specification, I would say that yes, 2 this is a proven design. Anyone who says that 3 there is a design defect, would have to come up 4 with some very specific rationale, scientific 5 rationale, to show why the 40-year history and 6 hundreds and millions of cylinders that have 7 been manufactured are all defective. 8 Q. So if I understood your answer, correctly then, 9 in your mind it's a combination of both the 10 history of performance of a particular product 11 in conjunction with whether it meets the design 12 specifications of CFR 178.65? 13 A. That's correct for this case. Your statement in 14 general, as a general statement, not specific to 15 this case, I would not say that specifications 16 by themselves are proof of an adequate design. 17 However, given the history that we have here, 18 you have to say those specifications have 19 certainly lasted the test of time. Someone had 20 better come up with a very specific design flaw 21 that is going to explain why the test of time 22 doesn't apply here. 23 Q. In a broader sense, if there would were to be, 24 for example, a new product that you had never 25 1 seen before, one that was, you know, first 2 manufactured this year. But objectively it met 3 all the design specifications of CFR 178.65, 4 would you be able to say, as a matter empirical 5 fact, that it has no design flaws without any 6 history to go on? 7 A. No. I couldn't. I think we are communicating 8 exactly on track. That is exactly what I'm 9 trying to say. If it was a brand knew design, 10 even though it meets the specs, I'd have to look 11 at the design. When it is an old proven design 12 with a substantial time history, then you really 13 have to come up with something very specific to 14 say it's not a good design. If it meets all the 15 specifications that have been around for 16 decades. 17 Q. So I'm trying to think of a way to shorten this, 18 so I don't have to keep looking back for the 19 reference number. Can I refer to CFR 178.65 as 20 DOT 39? 21 A. That's fine. 22 Q. For our purposes here? 23 A. For our purposes. 24 Q. Otherwise I will keep flipping back. DOT 39, 26 1 does that take into account the cylinder being 2 connected to a torch in terms of -- strike that. 3 To your knowledge, does DOT 39 take 4 into account the fact that the cylinder when it 5 had been used attached to a torch? 6 A. Not it doesn't. DOT is related to transport of 7 cylinders containing hazardous materials. Which 8 it means not in actual use, but in transport. 9 Q. Is DOT 39 applicable -- does it take into 10 account the type of material contained within 11 the cylinder? 12 A. It does to a certain extent. It specifies that 13 it should be a steel of a certain composition 14 and certain thickness. And it goes through some 15 other things. In terms of material, it says it 16 should be a steel of a certain composition 17 basically. 18 Q. Actually, I didn't ask my question very clear. 19 I apologize for that. Does DOT 39, it takes 20 into account the design specifications for a 21 cylinder that is intended to carry a volatile 22 gas; is that correct? 23 A. Yes. 24 Q. Is it volatile gas in the generic sense or is it 27 1 specifically for MAPP gas? 2 A. Well, it is for, I will say hazardous materials. 3 It doesn't even have to be a traditionally 4 flammable material, if I remember correctly, 5 because I believe I have worked with it on some 6 cylinders that were doing calibration gases for 7 the exhaust whether the checking the emission in 8 your cars. Basically it is gas cylinders going 9 over the highway. When those gases whether they 10 are flammable or not, have a hazard. It could 11 be a flammability hazard, it could be a toxicity 12 hazard; I would have to go back an review it. 13 It is not just flammable materials. 14 Q. A gas that has a hazardous quality to it, I 15 guess? 16 A. Yes. 17 Q. In your opinion, should the design of -- strike 18 that. 19 One more housekeeping issue. I'm 20 going to call these -- I guess I was going to 21 BernzOmatic one pound MAPP gas cylinders. The 22 problem is in this case, we don't know whether 23 the cylinder at issue was manufactured by either 24 Western or Worthington Industries. I'm going to 28 1 refer to it as BernzOmatic? 2 A. Why don't we just call them NRT cylinder; okay. 3 It wasn't manufactured by BernzOmatic in any 4 case. 5 Q. I understand that. We don't know who it was 6 manufactured by in this case, right. 7 A. But it is an NRT cylinder in any case. 8 Q. All right. Should the design of any NRT 9 cylinder take into account the intended use of 10 that cylinder? 11 A. Should the the -- I'm sorry. 12 (Record read as requested.) 13 A. Ultimately yes. Yes. 14 Q. In this case the design of the NRT cylinders 15 that are manufactured at the Chilton, Wisconsin 16 facility at one time by Western currently -- 17 more recently by Worthington take into account 18 the fact that these cylinders are intended to be 19 used with, for example, a TS 4000 torch. 20 A. Well, certainly the manufacturer of the cylinder 21 should be aware of the types of uses. So the TS 22 4000 torch, which has been around for a while is 23 something that you would expect that they would 24 know about. And if they have a problem with, 29 1 they should be notifying the manufacturer of 2 torch if there is a problem. By the same token, 3 new applications for these cylinders come along 4 all the time. Whether it is lanterns for 5 camping or cooking stoves or bug foggers. I 6 don't know that you use MAPP gas for bug 7 fogging. New applications come along all the 8 time. So I don't know that the cylinder 9 manufacturer has to be aware of everything that 10 anyone could go out there and use this for 11 indiscriminately. Okay. If that answers your 12 question. 13 Q. Are you aware of any standards take into account 14 or consider, in this case the NRT cylinder 15 manufactured at the West Chilton, Wisconsin 16 facility for BernzOmatic and the torch together? 17 A. I'm not aware of a specification that covers 18 that, a public specification by some society or 19 government. 20 Q. Have you reviewed the design specifications that 21 exist for the NRT cylinders that are 22 manufactured at the Chilton, Wisconsin facility? 23 A. I'm familiar with DOT 39. And I'm not sure what 24 else you are talking about. I have seen 30 1 drawings from Worthington for these, cylinders 2 which is part of their specific design. 3 Q. Okay. So you have seen the designs that -- for 4 the cylinder that -- the drawings, I should say, 5 that Worthington has? 6 A. Yes, I have seen some of the process 7 specifications for like the brazing process and 8 how the braze is applied and things like that. 9 Q. Have you ever visited the Chilton, Wisconsin 10 facility? 11 A. No, I have not. 12 Q. Have you ever -- strike that. 13 Have you ever seen a video or 14 photograph depiction of the manufacturing 15 facility and the equipment used? 16 A. I seem to remember in one case that I did see a 17 video that someone may have taken of the process 18 I can't for sure that the Chilton facility. But 19 certainly I do remember seeing cylinder 20 manufacturing facility videos. I have been 21 through cylinder manufacturing facilities, not 22 necessarily for NRTs and not necessarily Chilton 23 or Worthington, Westford. 24 Q. The videos you have a vague recollection of, I 31 1 don't mean to be judgmental, that you have a 2 vague recollection of seeing, do you recall 3 whether that was of the Chilton facility? 4 A. I seem to remember sometime in the past, that 5 there may have been in one of these other cases 6 that the plaintiffs may have had a tour of the 7 facility in the litigation case. I can't say it 8 was specifically the Chilton western facility. 9 Because I have seen such types of videos a 10 number of times. I also, like I say, I have 11 worked through myself other cylinder 12 manufacturing facilities and this goes back at 13 least 10 years. But it is vague, yes. 14 Q. Have you reviewed the drawings or designs that 15 BernzOmatic has for the TS 4000 torch? 16 A. I have seen them. I haven't reviewed them in 17 great detail. 18 Q. Are you able to give me an opinion as to what 19 area of the MAPP gas cylinder is under the 20 highest stress when it is connected to the TS 21 4000 torch during normal use? 22 A. I guess I'm having a little problem with what we 23 mean by normal use. If we are just talking 24 about internal pressure, then it's not at the 32 1 neck. If we are talking about a bending stress 2 due to a -- I will call it an eccentric load on 3 the torch on the nozzle then I do know where the 4 highest stress is. 5 Q. I was talking about the later situation. I'm 6 not talking about internal pressure. I'm 7 talking about TS 4000, the torch being attached 8 to the cylinder and the user handling in it an 9 intended manner such as soldering pipes. In 10 that scenario, where would the highest stress 11 load be? 12 A. The highest nonpressure stress, would be located 13 at the nozzle connection in the cylinder. 14 That's not necessarily the highest stress. 15 Q. It's not necessarily the highest stress? 16 A. That's correct. 17 Q. Because of different types of stress? 18 A. There is the internal pressure stress. What we 19 call the hoop stress in the cylinder, may be a 20 higher stress than what you have in the nozzle. 21 Q. But in terms of stress that comes from external 22 forces, the highest point would be the threaded 23 -- I don't know how you described it, the valve 24 housing unit? 33 1 A. If you get a bending moment on the length of the 2 nonreturnable tall cylinder and the torch, then 3 the highest stress location will be at the 4 nozzle. That is just inherently geometry. It 5 depends on the magnitude of that stress and 6 whether that stress is larger than the internal 7 pressure stresses or not; depends on the 8 magnitude of the bending. 9 Q. Are you aware of any public or governmental 10 standards that specifically refer to the nozzle 11 area of the NRT cylinder? 12 A. Not specifically to NRT cylinders. DOT 39 13 doesn't even specifically talk about the nozzle 14 area. They just talk about braze areas and such 15 things. Turns out the nozzle area is inherently 16 going to be a braze area in the design of one of 17 these cylinders whether it is NRT or another. 18 There are plenty of codes and standards that 19 talk about nozzle design and things like that. 20 Such as the ASME boil and pressure vessel code. 21 But doesn't applies to these cylinders. DOT 39 22 is what applies to these cylinders. 23 Q. ASME is that American Society of what? 24 A. Mechanical Engineers. 34 1 Q. Referring back to section three of your report, 2 the final sentence reads, "Thus, there is ample 3 evidence by physical test, by experience, by 4 calculation and historically, that there are no 5 design deficiencies in these fuel gas cylinders. 6 Can you describe for me the evidence 7 that you are referring to? 8 A. Well, the physical tests are the ones that are 9 required by DOT 39 and the others that are, in 10 addition, that are carried out by Worthington or 11 Western in the manufacture of these things. 12 They do design testing on individual cylinders 13 and then on every cylinder that is sold, they do 14 proof testing to ensure that each one has met 15 quality requirements. Experience, we are 16 talking about the types of failure rates that 17 have occurred over time. Calculations, there 18 are calculations that can be done on the service 19 stresses, the ability to handle pressure, the 20 physical testing that people have done to bend 21 or break these things. And historically we are 22 talking about, again, the kind of failure rate 23 and the overall experience over decades of use. 24 Q. In terms of the physical tests that are 35 1 conducted at the western Worthington, Wisconsin 2 facility, what is your source of information 3 with regard to the tests that they conduct? 4 A. Some of those tests are required by DOT 39. 5 They require a factor of two safety, if I 6 remember, in DOT 39. But in fact I have seen in 7 documents, I believe probably documents that 8 have been supplied in this case but burst 9 pressure tests that are done on the 10 manufacturer's cylinders and then after -- 11 periodically they will burst some of these 12 cylinders. And typically on an NRT cylinder, 13 you have a burst pressure of around a thousand 14 PSI is the failure pressure for this shaped 15 cylinder and size cylinder. There are other 16 tests, such as they -- after manufacture they 17 check for leaks by putting these in a bath of 18 hot water. I think Mr. Gentry had talked about 19 the fact that the maximum design pressure is 188 20 PSI, but each cylinder would get something -- be 21 tested at something like I think is numbers 319 22 PSI. But in fact the breaking pressure is on 23 the order of 1000 PSIs. That's where you get 24 these -- if you read this whole paragraph three 36 1 where, I say, The standard requires the safety 2 factor of two, but in fact the actual safety 3 factor on this NRT design is actually closer to 4 a factor of 5. Safety factors are greater than 5 five is what I said. 6 Q. In the earlier part of your answer, you made 7 reference to some documents. I will ask you 8 about your file in a moment. Were you referring 9 to the deposition testimony that Steven Gentry 10 gave in this case with regard to his description 11 of the tests that were conducted? 12 A. Yes, I read his deposition. 13 Q. I wanted to know if there are other documents 14 related to the testing procedures and protocols 15 that are used at the Chilton facility besides 16 Mr. Gentry's deposition? 17 A. I'm not certainly relying on any for this case. 18 I believe they were appended to his deposition 19 so far as that goes. Certainly I have seen 20 documents at other times. But for the MAPP gas 21 cylinders it would have been either this case or 22 the Glenn case. I don't think I read very much 23 on him or if even he gave a deposition in the 24 Sandoval case. I have read things that he has 37 1 talked about but generally I haven't seen 2 anything inconsistent between depositions in 3 different cases that he has given. 4 Q. Okay. Have you brought your file with you 5 related to this case? 6 A. Yes. 7 Q. Maybe we could go off the record a moment. I 8 would like to flip through it and identify 9 anything I feel needs to be identified? 10 (Break in the proceedings.) 11 MR. EPSTEIN: Back on the record. 12 BY MR. EPSTEIN (CONT'D): 13 Q. I want to get some housekeeping matters out of 14 the way. Dr. Eagar, you have given what is 15 essentially your file for this case; is that 16 right? 17 A. That's correct. 18 Q. I'm going to identify the bulk of it by category 19 not specifically item by item. There is a just 20 few things I will ask you questions on in a bit. 21 One of the, I will call subfiles that you've 22 handed me appears to be copies of deposition 23 transcripts and a deposition notice in this 24 case. It appears mostly to be copies of 38 1 deposition transcripts from this case; is that 2 what this file is? 3 A. That's correct. 4 Q. I'm going to go ahead and give this back to you. 5 Did you read all of those deposition 6 transcripts? 7 A. Yes, I did. 8 Q. Another big bulk, rubber-banded together appears 9 to be what was Exhibit 6 is Dr. Anderson's 10 deposition last week in this case, is that what 11 your understanding of what this rubber banded 12 section is? 13 A. Yes. 14 Q. Did you review Exhibit 6 to Dr. Anderson's 15 deposition? 16 A. Yes, I did. 17 Q. We have a number of CD ROMs and I think a DVD. 18 Are these CDs and DVDs, do they contain still 19 photographs that you understood came from Dr. 20 Anderson? 21 A. Some of them do. Most of them probably do. I 22 think one of CDs is actually a professional 23 video made of Mount Shasta testing. 24 Q. Is there anything else to your knowledge, 39 1 depicted on any of these disks? 2 A. Not that I remember, no. 3 Q. These disk were supplied to you by Mr. Ergo's 4 office? 5 A. Yes. 6 Q. I will go ahead and return these to you: We 7 have a copy of the stipulated protective order 8 in this case. I will go ahead and give that 9 back to you. 10 A. Okay. 11 Q. We have a plastic protective covering of a copy 12 of what appears to be Dr. Anderson's June 25, 13 2008 report along with Plaintiffs' disclosure of 14 expert -- liability expert reports? 15 A. I think you will find Dr. Vrendenburgh's report 16 in there somewhere. It is basically the 17 Plaintiffs' report. 18 Q. In looking at your copy of Dr. Anderson's June 19 25, 2008 report, you have highlighted in yellow 20 the caption on page six above a photograph. The 21 caption reads Photo to close up of Gellen v 22 Newell cylinder at the valve housing showing a 23 brazing material failure. You see you have 24 highlighted the words showing a brazing material 40 1 failure? 2 A. Yes. 3 Q. Can you tell me why you highlighted that portion 4 of the caption? 5 A. Because that is not a brazing material failure. 6 Q. You disagree with a Dr. Anderson's opinion that 7 is a brazing material failure? 8 A. That's correct. 9 Q. Why do you disagree with that? 10 A. I have inspected the cylinder myself. It is a 11 cylinder wall steel failure. 12 Q. Okay. Is it a steel failure in the area of the 13 joint between the valve housing unit and the 14 cylinder wall? 15 A. Yes. But it is in the steel not the braze. 16 Q. The Glenn versus Newell cylinder, that cylinder 17 was at some point sectioned, was it not, to your 18 knowledge? 19 A. Yes, it was. 20 Q. Was parts of it examined under a microscope? 21 A. Yes. 22 Q. Were you present when that was done? 23 A. Yes. 24 Q. Did you examine a portion of the brazed joint 41 1 under the microscope? 2 A. Yes, I did. 3 Q. Can you tell me, were you able to make a 4 determination in your mind as to whether there 5 was any failure of the brazing material? 6 A. There was a fill-it of the braze or meniscus, as 7 Dr. Anderson calls it, that the meniscus had 8 separated but it was because the structural 9 failure was through the steel. 10 Q. Which in turn caused the meniscus to separate? 11 A. Well, the meniscus was on top of the steel 12 fractured. The highest stress location in the 13 steel has a meniscus over it. And in order to 14 fracture the steel, which is substantially 15 stronger than that meniscus it also fractured 16 the meniscus in two. 17 Q. Did you come to an opinion as to what caused the 18 steel to fracture? 19 A. The torch had been bent to the side at an angel 20 of approximately 20 degrees or 25 degrees. We 21 measured it. I would have to go back and look 22 at the photos. It was more than 15 degrees, it 23 was 20 or 25 degrees. 24 Q. What does that signify? 42 1 A. Well, the metal had been deformed and it shows 2 that it been abused. It takes by Dr. Anderson's 3 reckoning, his measurements, about 30 foot 4 pounds to do that. 30 foot pounds by Dr. 5 Anderson's report is beyond the normal use 6 range, expected range of forces. It is sort of 7 like banging and using the torch and cylinder 8 like a hammer, that is hard someone had to hit 9 that in order to get that deformation in 10 bending. 11 Q. Could that deformation or bending have been the 12 result of a force that emanated from within the 13 cylinder? 14 A. No. 15 Q. Why not? 16 A. Because you have had a failure -- the highest 17 stresses from within the cylinder, being the 18 pressure stresses, would be a hoop stress, which 19 will cause a split axially along the side of the 20 cylinder. The stress is on the head of the 21 cylinder where the nozzle is are approximately 22 half the stress value of the axial hoop 23 stresses. So that we know this is the low 24 stress location from an internal pressure point 43 1 of view. That is just simple geometry. 2 Q. You also highlighted a section on page three of 3 Dr. Anderson's report. A portion of the 4 sentence which reads, From the reference, quote, 5 Hansen, H-A-N-S-E-N, constitution of binary 6 alloys, closed quotes, that temperature is too 7 low to melt the brazing alloy. Can you tell me 8 why you highlighted that portion of the 9 sentence? 10 A. Because he's using a handbook of binary alloys 11 when in fact this is a ternary alloy that 12 contains phosphorus and even his measurements 13 which are appended later in the report show that 14 there was four percent phosphorus found in the 15 alloy. Therefore by ignoring the true 16 composition he comes to conclusions about the 17 melting point temperature that are just science 18 fiction. 19 Q. The Hansen constitution of binary alloys 20 reference is that not an appropriate reference 21 for purposes of the analysis being conducted 22 here? 23 A. You can't get the melting temperature of a 24 ternary alloy by looking at a binary diagram. 44 1 You are leaving out one-third of the 2 information. 3 Q. Below that in his section entitled conclusions, 4 slash, findings, you've highlighted a portion of 5 a paragraph, first sentence you highlighted 6 reads, The brazing materials have large voids in 7 the bulk and smaller voids in the interface 8 between the cylinder wall and the center valve 9 housing as shown in microphotographs six, seven 10 and eight. Can you tell me why you highlighted 11 that sentence? Or should I read the rest of it? 12 A. No. It is just something to -- when I go back 13 and look at his report, look at the issues that 14 he is raising. 15 Q. You have also highlighted a portion of the 16 paragraph that reads, The brazing on the outer 17 surface of the cylinder is undercut in all three 18 cylinders rather forming a meniscus. The 19 undercutting is a sign of lack of wetting and 20 penetration of the brazing material with the 21 cylinder valve. 22 Is there anything more of significance 23 to that other than to highlight it so you can go 24 back? 45 1 A. It tells me what his opinion is. I'm 2 highlighting it, because it is based on false 3 assumptions. 4 Q. What are those false assumptions? 5 A. He doesn't understand how the braze allow is 6 placed on the joint prior to heating it in the 7 furnace. And no one would ever draw the 8 conclusions that he draws there if they 9 understood how the braze alloyed flowed in 10 furnace from where it was initially placed. 11 Again, he is guessing at the manufacturing 12 process and he is guessing wrong. 13 Q. How do you know he is guessing at the 14 manufacturing process? 15 A. Because you could never draw that conclusion 16 that the metal didn't flow if you knew where it 17 started and where it ended. He knows where it 18 ended. But he doesn't know where it started. 19 He said it didn't flow properly. Well, if you 20 knew where it started and you knew where it 21 ended. You know the path it had to take to get 22 there. He knows one end of that path. He is 23 drawing a conclusion of what the entire path is. 24 But by knowing only the end of the path, you 46 1 can't draw a conclusion about whether the path 2 was an appropriate path or not. He has no idea 3 what the true path is. 4 Q. With regard to the start of the path, was that 5 described in the deposition testimony of Mr. 6 Gentry? 7 A. I don't remember if it came up. It is actually 8 in a document that I have seen, Mr. Gentry 9 showed it to me showing where the braze alloy is 10 placed. I specifically asked him that question. 11 Q. Is that in your file? 12 A. No, I don't believe I have that document. 13 Q. Did you rely on that document in forming the 14 opinion that you just expressed regarding Dr. 15 Anderson's opinion? 16 A. Yes, I did. 17 MR. EPSTEIN: I would like to request 18 a copy of that document. 19 MR. ERGO: It shouldn't be a problem. 20 Follow up with me on it, though. I will get on 21 the plane and forget about it, if you don't. 22 MR. EPSTEIN: Okay. 23 Q. Dr. Eagar, what is the document that you are 24 referring to, if it isn't in your file, tell you 47 1 with regard to the placement of the brazing 2 material? What does it tell you? 3 A. If you look at the nozzle and the little ring of 4 where the meniscus is on the outside, the braze 5 alloy is placed at four clock positions; we will 6 call them 12, 3, 6, and 9. They have kind of 7 like a little syringe type of the tool that 8 comes in and squirts some of the braze alloy 9 paste at those four positions. Then you put it 10 in the furnace and it flows from those four 11 locations down through the joint to the bottom 12 of the joint. It would be impossible to not 13 have flow, proper flow and penetration of the 14 joint if it starts at those positions and flows 15 all the way around from 12:00 through 3, through 16 6, through 9, through 12. That is one of the 17 quality control inspection techniques that is 18 used in the Chilton facility is to look and see 19 if you get braze all the way around that is 20 connected between 12 and 3, and between 3 and 6 21 and 6 and 9 and and 9 and 12. If it connected, 22 you know in flowed. 23 In addition, we know from the cross- 24 sections that Dr. Anderson had made at Anomet, 48 1 that the braze flowed all the way down around 2 the corner to the bottom of the joint. Because 3 there was never any braze placed at the bottom 4 of the joint. So if you know where it started 5 at the top in four different clock positions, 6 and you know that afterwards, by looking at it, 7 that it flowed all the way around 360 degrees at 8 the top, and you know that it flowed all the way 9 down to the bottom, you can't draw the 10 conclusion that he draws. It is impossible. 11 That is unscientific. 12 Q. And you also have highlighted a section closer 13 to the conclusion of Dr. Anderson's report. I 14 will read the entire sentence even though it is 15 only partially highlighted? 16 MR. ERGO: Which page? 17 MR. EPSTEIN: Four. 18 Q. It reads, "Corrosion test show that the brazing 19 material is strongly cathodic to the cylinder 20 and valve and will cause the steel to corrode in 21 a suitable moist atmosphere." 22 Can you tell me why you highlighted 23 that portion? 24 A. Because at the time I read this, I had seen no 49 1 evidence of that testing. And, in fact, I think 2 that's the -- one of the disks that was sent, I 3 think I contacted Mr. Ergo and said, Hey, we 4 don't have any -- he did some testing on 5 corrosion. We don't know anything about it. 6 His report just says he did corrosion test. But 7 we don't know how he did it. And I actually did 8 dispute at the time whether that conclusion, 9 okay. Now that I know how he did it, I think 10 that what he describes there he has 11 misrepresented what he did. 12 Q. Why do you feel he has misrepresented what he 13 did? 14 A. Because he immersed in a three percent salt 15 solution, which is basically seawater. I 16 wouldn't call seawater an appropriate, moist 17 atmosphere. 18 Q. Why not? You consider it more than? 19 A. I consider. 20 Q. More than moist atmosphere? 21 A. I consider storing one of these cylinders 22 immersed in seawater for two days to be an 23 inappropriate test and not representative of 24 whether a moist atmosphere, such as this humid 50 1 room that you pointed out earlier today off the 2 record, which is where these cylinders are 3 stored typically. I don't think that test of 4 immersing it in basically a salt solution which 5 is almost as salty as seawater is anywhere close 6 to being appropriate to the type of moist humid 7 atmosphere that you expect these cylinder to 8 undergo in a normal service. I don't think 9 people should be storing these things in the 10 ocean. I don't think that test is an 11 appropriate test. I don't that if anyone goes 12 and looks at cylinders that have been stored in 13 a cabinet or at Home Depot. I have a cylinder 14 in my office, which I looked at it yesterday, it 15 has a B on it, which means it was manufactured 16 in 2002, it spent most of its life in a cabinet 17 in my laboratory since 2002. You look at that 18 and you see absolutely no corrosion. As you can 19 tell Cambridge, Boston area, we are close to the 20 seacoast, we get humid. We have lots of trees 21 around here. And there is no corrosion after 22 six years. For him to start suggestion that 23 this is a bad design because of corrosion goes 24 and flies in the face of 40 years and hundreds 51 1 of millions of cylinders. It just shows to me 2 an advocacy position that he is taking here. 3 Because when you read his deposition, he 4 admitted that can't relate the corrosion to 5 anything that had to do with failure mode. Why 6 is he even bringing it up? He is just trying to 7 bad mouth these cylinders any way he can. 8 Whether it relates the -- he admits it doesn't 9 relate to the failure of the Shalaby cylinder. 10 That is what he admitted in his deposition. Why 11 is it in his report. It is irrelevant. 12 Q. Before I'm going to finish up my inventory on 13 the record of your file. A couple of things -- 14 strike that. 15 The photographs -- the copy of the 16 photographs that you have stored with Dr. 17 Anderson report and Ms. Vrendenburgh's report 18 have a couple, two or three orange stickies on 19 them. One photograph in particular, you have an 20 orange Post-it in which you have written start 21 of leak and another Post-it in which you've 22 written angle 59 degrees, bend 59-27 equals 32 23 degrees. The information that you have written 24 on these orange tabs, did that play a part in 52 1 any of the opinions expressed in either your 2 report or supplemental report? 3 A. Well, I didn't have those until after my 4 original report. So they didn't apply to my 5 original report. And I don't remember if we 6 actually had taken those stills off the 7 professional CD video prior to my second report 8 or not. I'm pretty sure I had seen it. I don't 9 remember if we had gotten the measurements at 10 that point. But I could tell from the video 11 when I saw it in July, probably prior to, that 12 there was a significant bend up at Mount Shasta 13 when they put the socket wrench on the head and 14 started pulling on it. I could tell there was a 15 significant bend. I had not quantified it at 16 that time, but I could tell there was a 17 significant bend. The purpose of those 18 photographs, we took the stills off the video. 19 Then we measured from the stills what the bend 20 angle was. A 27 degree bend angle is a 21 substantial bend angle prior to getting leaking. 22 In the Glenn case, we got leaking at about 20 or 23 25 degrees. Anything above five or ten degrees 24 is abuse. 53 1 Q. Okay. I'm going to hand this subfile back to 2 you. Also I have one other subfile with plastic 3 covering. It looks like it contains in addition 4 to our disclosure of expert reports, Plaintiff's 5 disclosure. I can't quite tell. It looks like 6 a copy of the reports from the experts which 7 BernzOmatic or Irwin Industrial Tool Company 8 retained, is that what these are? 9 A. Probably. I probably received them in two 10 different e-mails and printed them out a day 11 after, one after the other. 12 Q. Did you review the reports of -- the report and 13 supplemental reports of Dr. Kane and Myers? 14 A. Yes, I did. 15 Q. Have you formed any opinions regarding the 16 opinions they expressed in their reports? 17 A. I would have to go back and look. I haven't 18 looked at it for a while. Yes, I do have at 19 least one opinion. 20 Q. What is that opinion? 21 A. Well, in the report of Dr. Myers and Dr. Kane, 22 they say that it is not possible to fully test 23 or analyze any of these hypotheses or to 24 determine the cause of the incident. And I 54 1 disagree with that and I think my whole first 2 report shows that you can eliminate design 3 manufacturing or materials as a probable cause 4 of this incident. And I think that there is 5 substantial testimonial evidence; however much 6 you want to rely on it, it is more than just one 7 person said X. Okay. That there may have been 8 abuse. There also is the testing that has been 9 done and the physical evidence that was 10 testified to by the park rangers that there was 11 a split where they describe the split was not at 12 a braze -- could not have been at a braze joint. 13 It was too long to be at the braze joint. To be 14 that long, you would have to go almost all the 15 way 360 degrees circumferentially around the 16 braze. We know that the braze joint is 17 substantially stronger than the sheet steel 18 itself. There is no reason ever to expect a 19 failure in the braze even if it has porosity. 20 We also know that one of the rangers described a 21 bend or tilt to the torch, I think he said five 22 or ten degrees. But the size split is not 23 consistent with a five or ten degree bend. It 24 might be more consistent with a 20 or 25 degree 55 1 bend. But in any case, there is substantial 2 evidence that is indicative of abuse. So in my 3 opinion it is possible to determine that the 4 cause of this failure was abuse. Now whether it 5 was kicking it into the fire, banging it against 6 the fire ring or hitting it with a sledgehammer 7 I wasn't there. I can't tell you which specific 8 abuse it was. But I can tell you, you will not 9 get those physical artifacts described by the 10 park ranger. 11 I will tell you that I don't believe 12 that Mr. Shalaby would be making some of the 13 statements that he made immediately afterwards 14 that were adverse to his interest or that other 15 people would have been making some statements, 16 multiple people making statements, unless there 17 was the likelihood of some sort of abuse. Had 18 both the description of the evidence before it 19 was thrown out and I have testimonial evidence 20 of Mr. Shalaby and other people that indicated 21 that there was some, I guess if I want to borrow 22 a word, bazaar behavior going on. Inappropriate 23 behavior is maybe what I would say in the use of 24 this cylinder. I would call that -- put all 56 1 that together and say the weight of that 2 evidence, to me, suggests that it is possible to 3 say that this cylinder was abused. 4 Q. Is it your opinion in this case, that the 5 cylinder was in fact abused? 6 A. That is what the conclusion that I would draw 7 from the evidence -- the weight of the evidence. 8 I'm not relying on any one statement of one 9 witness. I take all these statements together. 10 I have half a dozen, ten statements depending 11 how many you want to count. Who is just relying 12 on someone else's statement. There is a half a 13 dozen to ten descriptions of physical evidence 14 or description of behavior that indicate this 15 was abused. I'm also relying on my knowledge of 16 the engineering analysis of the cylinder and how 17 it can fail. 18 Q. You stated a moment ago that there is no reason 19 ever to expect a breach or break in the braze 20 material even if it has porosity, can you 21 explain why that is the case? 22 A. The braze joint is at least twice if not two and 23 a half three times as long as required by DOT 24 39. DOT 39 requires that the braze joint have a 57 1 length four times the thickness of the sheet 2 metal. 3 Now, that is not an uncommon type of 4 specification in brazing of sheet metal, just 5 generally, to require that the brazed joint be 6 stronger than the underlying sheet metal. Okay? 7 It is an overlap joint is what we call it. If 8 the overlap is four times the thickness of the 9 sheet and you know that the braze alloy, I'm 10 talking any braze alloy, and in fact the DOT it 11 would be pure copper, not a copper alloy, but 12 pure copper which is clearly softer than steel. 13 Even if the copper is softer than steel, if 14 you've got four times the area of the braze as 15 you have of the cross-sectional area of the 16 steel, the steel is always going to lose that 17 stress match. If you are talking a tug-of-war, 18 between the steel and the copper. The copper 19 will always win, because it has four times the 20 area. Sort of like a tug-a-rope, if you want. 21 And you've got one person pulling for steel on 22 one side and four people pulling for copper on 23 the other. You have got four times the area and 24 the odds are cooper is going to win that 58 1 tug-of-war. 2 Well, it turns out in this case, we 3 had an alloy braze, which is actually closer in 4 strength to the steel than the Cooper would be. 5 We also have, if you go from the outside where 6 these little four drops of solder are placed and 7 you let the braze flow all the way down and 8 around to the bottom, you have an area that is 9 substantially greater than four times the 10 thickness. It is more like 10 times the 11 thickness or maybe even closer to 12 times the 12 thickness. 13 As a result, I have estimated that the 14 braze -- a perfect braze joint with no voids, 15 would be probably at least 30 times stronger 16 than the steel sheet metal. Even a braze joint 17 that has some voids, such as described in Dr. 18 Anderson's report, may only be 20 times stronger 19 than sheet metal. Now, if we want to consider 20 the geometry as well as the strength of the 21 material, we are talking about a tug-of-war 22 against one steel person and 20 braze people. 23 Who is going to win that tug-of-war. I 24 guarantee steel always loses. You will always 59 1 have the failure through the steel. When these 2 things have been failed at room temperature, I 3 have never seen a failure anywhere by bending 4 around this neck that is not in the steel. And 5 that is just scientific analysis. I have put 6 numbers on it. Okay. I have calculated some of 7 these things. As Dr. Anderson said, he could 8 put numbers on it, if had done it, at the time 9 of his deposition, he said he could. I think if 10 you are going to opine, you ought to put some 11 rough engineering numbers on it. You don't have 12 to put very precise numbers on it to say it is 13 about 20 times stronger. So why would I every 14 expect a failure in the braze. There is 15 absolutely no scientific evidence in this case 16 that we had a braze failure in spite of what Dr. 17 Anderson says. 18 Q. Very quickly in finishing up your file here. 19 Are these four invoices the sum total of the 20 invoices that you submitted for the work that 21 you conducted in connection with this case? 22 A. Yes, they are. 23 Q. I have got a clipped section of loose yellow 24 sheets, are these all the notes that you've 60 1 taken in this case? 2 A. When reading the deposition, that's correct. 3 It's actually all my notes. They happen to be 4 depositions because there is no inspection of 5 material or anything, there is no material 6 evidence. 7 Q. These are all of the notes that you have in this 8 case -- other than what might be written on 9 other pleadings and reports are within this 10 packet right here? 11 A. Right. You notice that basically I highlight 12 some of these other reports and such things. I 13 will tell you that I don't highlight the 14 depositions. Those are clean. I never mark on 15 the deposition, so I don't have to carry them 16 across the country with me. 17 Q. With regard to the top pages are your notes 18 regarding the deposition of Dr. Anderson that 19 were taken last week; is that correct? 20 A. That's correct. 21 Q. Some of the lines you have written asterisks 22 next to what you have written, what do the 23 asterisks represent? 24 A. Those are things that I consider key points. 61 1 Basically, I just write down all the lines and 2 by line number are just things is my summary of 3 the deposition. Anything that I think might be 4 important. A lot of times one of the first 5 things in a deposition will be who the person 6 was, where they worked, which is not really 7 critical to testimony. But things that I think 8 are significant from more of a scientific or 9 evidentiary point of view I will often put an 10 asterisks by. Sometimes if I think it is really 11 important, I will put two asterisks or three, or 12 if I really excited. 13 Q. What about the lines that appear to have numeral 14 next to one, vertical lines? 15 A. As you have pointed out I have done quite a few 16 depositions. There is unique. I have never 17 done this before. But Dr. Anderson has made so 18 many fundamental scientific errors, I started 19 keeping track of them, numbering them. Just 20 like you do chalk 1, 2, 3, 4 and a line for 5. 21 I actually decided, I didn't do it in the 22 beginning. I decided to go back and start 23 numbering the number of fundamental scientific 24 errors that he made. I mean, I have seen other 62 1 people make scientific errors. But he is almost 2 in a class by himself for how many false 3 assumptions that he then goes off, you know, 4 drawing conclusions from false assumptions. I 5 have never come across anything quite so 6 dramatic before. 7 Q. The first line where you have written a one 8 maybe with a closed parentheses? 9 A. No, he got two on that page. Two on one page. 10 Q. Your notes appear to read, page 34 thinks 11 cooling is -- 12 A. Is the Joule-Thomson effect. 13 Q. Of the MAPP smells due to mercaptans? 14 A. Mercaptans, M-E-R-C-A-P-T-A-N-S. 15 Q. You found two fundamental errors in his analysis 16 there? 17 A. Yes. They are not related to the failure at 18 all. It just shows the unscientific nature of 19 his analysis. The cooling of the gas in the 20 cylinder is not due to the Joule-Thomson effect. 21 It is due to the fact that you are taking the 22 vapor off and leaving the liquid behind. So 23 every bit of gas you take off is caring with it 24 what we call the heat of vaporization. 63 1 Now, I don't fault him too much for 2 this, because one of my colleagues at MIT when I 3 was doing a clam bake once and he saw the 4 propane tanks cooling down, he said, "Oh, that's 5 the Joule-Thomson effect." I said, "No, Kent 6 that's heat of vaporization. He walked away 7 sort of embarrassed." 8 But nonetheless, other people have 9 made that mistake but it shows a lack of 10 thermodynamic understanding of what is going on 11 with the gas. 12 Q. What is the other error in his testimony of page 13 34? 14 A. He basically indicates that the smell is due to 15 mercaptans. There are no mercaptans in MAPP 16 gas. The mercaptans are put into natural gas, 17 which is primarily methane. Because methane 18 being a nonpolar molecule doesn't have much of a 19 smell. They actually introduce what they call 20 an odorant, which is a sulfur compound, which is 21 a mercaptan into the distribution gas in your 22 homes, that goes out to your homes. It doesn't 23 come from Louisiana that way in the big trunk 24 lines. Once it gets to Boston, NSTAR, is the 64 1 company around here, they put mercaptans in it 2 so that if you get a leak, you will smell it. 3 Now, MAPP on the other hand is 4 nonpolar molecule and it has a smell all of its 5 own. Actually acetylene, which is mostly 6 nonpolar, mostly nonpolar or actually has a 7 smell. But basically MAPP gas has its own 8 smell, you don't need to put odorants in. It 9 shows his unfamiliarity with MAPP gas. 10 Q. Then you have got another Roman numeral next to 11 a line that reads page 52, braze joint is 12 weakest based on his testing? 13 A. Right. 14 Q. What is that -- can you explain the fundamental 15 error that he has made? 16 A. I have already explained how the brazed joint is 17 even with voids as he has seen is not the 18 weakest. It is actually probably 20 times 19 stronger than the steel. And he says it is 20 based on his testing. But he shows actually no 21 test that shows a failure in a braze joint 22 except the one at Mount Shasta that he heated up 23 before Mr. Snyder, Chris Snyder, basically 24 reached over and took the hot thing and pulled 65 1 on the wrench in a different direction and 2 opened it up more. There is a braise failure 3 there. But it was after he had been cooking in 4 a big fire of MAPP gas for about four or five 5 seconds. The only time they got a braise joint 6 failure is when it was being heated to melting. 7 There has never been, in my knowledge, in any 8 case I have ever been involved in a failure of 9 braze joint. That is because the braze joint is 10 20 times stronger than the sheet steel that it 11 is joining together. 12 We are talking now somewhere between 13 50 and 100 cylinders that I've looked at, 14 whether they are MAPP or propane or anything, 15 with these NRT type of cylinders. So he 16 concluded the Glenn cylinder, which he hasn't 17 even inspected was a braise failure. He said it 18 in his deposition unequivocally. Well, I 19 inspected it. Dr. Kane and Dr. Myers inspected 20 it. We came to the conclusion it was a failure 21 in the steel. Anyone who looks at it and looks 22 at the cross section, knows it is a failure in 23 the steel. But Dr. Anderson looks at a photo 24 and says, Oh, it is a braze failure. There is 66 1 about as unscientific as I have never seen. 2 From someone who has the type of qualifications 3 that he has. 4 Q. I'm sorry, you mentioned the significance of 5 asterisks before is that different than the 6 Roman numerals? 7 A. Right. The Roman numerals -- some of the Roman 8 numerals, Joule-Thomson and the mercaptans, has 9 nothing to do with the failure. It just shows a 10 lack of knowledge about the whole system. It 11 shows a lack of expertise on MAPP gas and 12 cylinders and thermodynamics and gases and how 13 they escape. The other things with the 14 asterisks should relate more specifically to the 15 Shalaby case and how the accident occurred. 16 Q. Okay. The line -- it says page 116 Shalaby 17 failed due to 12 to 30 pounds on tip. Why did 18 you asterisk that line? 19 A. Well, that actually -- that number comes from 20 the 30 foot pounds in Dr. Anderson's report. In 21 addition to in his deposition where he is now 22 relying on some of the Dr. Carr's tests, which 23 we haven't had a chance to evaluate. Which I 24 don't think is exactly fair. If he is going to 67 1 start relying on Dr. Carr's test, are you going 2 to let us depose Dr. Carr. Let me know what Dr. 3 Carr did before we start relying on him. I will 4 take the 30 foot pounds. I wrote down because 5 that's what he said. If we take the 30 foot 6 pound, first of all, he is being a little 7 imprecise, because it is not 30 pounds on the 8 tip, it 30 foot pounds. The tip is only 10 9 inches away, it is not 12 inches away. It's not 10 a big deal. But it really would be 36 pounds on 11 the tip. That's what he said. That's just to 12 remind me what he said. When I corrected it, it 13 should be 36 pounds on the tip based on his 14 test. If you want to go to Dr. Carr's test then 15 we would be talking about 14 pounds on the tip. 16 But I don't want to Dr. Carr's test unless I can 17 evaluation Dr. Carr and how he did it. I don't 18 know if it is the same test. 19 Q. Where it reads page 121 get to the 12 to 30 20 pounds with dynamic forces. 21 A. Yes. That actually 121 and 124 go together. 22 There is where he is claiming that you can wave 23 it around in the air. You can have the torch on 24 it, and hold it by the torch and wave the tank 68 1 in the air. Well, if that is true, if that 2 physics is true, you'd better not turn any 3 corners in your car. Because I will you, you 4 have gas in the tank of your car. If the liquid 5 sloshing in your gas tank is going to rip that 6 tank apart, we are going to have a lot of people 7 blowing up because of gasoline fires as we turn 8 the corner. 9 He said he hadn't done the analysis, 10 he said he bet he could do it. As I was reading 11 his deposition, I did the analysis. In fact, I 12 did it the hard way. I actually went back 13 through my head and ended up calculating that if 14 you had a full tank, which would be the worst 15 case, which weighed about two pounds and the 16 center of gravity is five pounds below the 17 nozzle, then you actually would have to have, I 18 calculated it by doing it in my head about 75 19 pounds of force due to this sloshing of the gas 20 or liquid in the tank. 21 Now, then I realized when I went all 22 the way through the hard way to do the 23 calculation in my head, I realized that is the 24 same as the 36 pounds divided by two, which is 69 1 72, multiplied by two, divided by half. Because 2 instead of 10 inches at the tip. I have 72 -- 3 instead of 36 pounds at a 10-inch tip, I have 72 4 pounds at a five inch distance to the center of 5 gravity. Okay. All I had to do is 36 by .5 and 6 I have come up with 72. Now, doing it in my 7 head and rounding off, I come up with 75. Then 8 I said, Oops. It is the same physics. Okay. 9 There is a shortcut to doing all of this. You 10 would still have to put 72 pounds of force on 11 center of gravity. 72 pounds of force on a 12 center of gravity is not just hitting it with a 13 hammer, it is taking both hands and running up 14 to the table and slamming it against the table. 15 Now if he thinks he can do that by just waving 16 it in the air with your wrist, I don't think so. 17 I would like to see his calculation. He said he 18 could calculate it. I would like to see that 19 calculation. I bet it is just as accurate as 20 some of the rest of his science. 21 Q. Okay. You have written a few lines down from 22 that with an asterisk next to it, page 139, if I 23 make the brazing materials really strong, than 24 what I am going to do is fail the cylinder? 70 1 A. Right. I agree 100 percent with that. Okay. 2 Q. The next line, page 140, that is going to take 3 most of the abuse -- that is going to take most 4 of the abuse that would be given to it, with an 5 asterisk next to it? 6 A. Right. 7 Q. Why did you asterisk that? 8 A. Because I thought that is interesting because he 9 is now admitting that if you get these types of 10 loads on it, that you have abused the cylinder. 11 In fact that's what he said in his report that 12 30 foot pounds was beyond the normal use of the 13 cylinder, expected use of the cylinder. So if 14 you really read what he says carefully, you are 15 finding that he admits that the cylinder was 16 abused in order to get this fracture. 17 Q. Down from that, you've written page 147, he 18 discounts the description of the metal curling 19 out since it is not consistent with failure of 20 the braze? 21 A. Yeah. That is sort of a fact. People were -- I 22 think Mr. Ergo as asking about this curling, 23 maybe it was someone else asking about the 24 curling, maybe it is Ms. Naylor. Anyway, 71 1 frankly, the problem I have with this curling 2 and explosion as I've already pointed out. The 3 stress is on the head of the cylinder due to 4 internal pressure or one half the stresses along 5 the axis of the cylinder. We know that from 6 simple geometry. It is taught to sophomores in 7 mechanics courses. So I just can't believe this 8 is an internal pressure failure of the nozzle 9 location. You can burst these things any way 10 you want. In fact, the British report that he 11 appends to his report shows that you can abuse 12 these all kinds of ways and they never could get 13 a failure of that nozzle. The way you can get a 14 failure of the nozzle is if you take it and bend 15 it or hit it against the table like you were 16 hitting a nail into a two-by-four. If you do 17 that, which I consider to be abusive, you can 18 get the types of failures that we have here. I 19 will not say that this cylinder is 20 indestructible. But I will say that the only 21 way you can get the failure that we have 22 described by the park rangers in the Shalaby 23 case is by abuse. That is what my first report 24 says. 72 1 Q. A few lines down, excuse me, you've written page 2 161, braze alloy is CU-NI -- 3 A. Copper, nickle, binary. We have already 4 discussed that already, I think, when you were 5 reading, his report, page three. 6 Q. Next line down, you have a Roman numeral one 7 next to that, page 165 claims that fusion 8 welding pressure could lower the melting point. 9 Then you have in brackets, no with an 10 exclamation point? 11 A. Right. 12 Q. What does that mean? 13 A. Basically that violates LeChatlier's Principle 14 of Thermodynamics. 15 Q. I can't say I'm familiar with that? 16 A. That basically says if you push on something, it 17 is going to respond in the direction you are 18 pushing, not in the opposite direction. 19 Q. Okay? 20 A. Basically if you add pressure to something that 21 expands on melting, then you are going to make 22 it want to solidify, not want to melt. He says 23 that by adding pressure, you will make it melt. 24 In fact LeChatlier's principle tells me from 73 1 basic thermodynamics that if I push on a 2 material at some temperature that expands on 3 melting that I'm going to raise the melting 4 point not lower it. He has got his science 5 reversed on basic physics. Now what he also 6 doesn't have is the fact that the pressure, you 7 would have to apply, just to get a temperature 8 change as big we are talking about, would only 9 be obtained somewhere a few hundred miles 10 beneath the surface of the earth. We would be 11 talking about somewhere in the hundreds or 12 thousands of atmospheres of pressure in to get 13 this type of differential. Anyone who has ever 14 done any thermodynamic calculations of the 15 affect of melting on temperature. In fact, it 16 is called the Clausisus Clapeyron equation that 17 gives you the temperature change with pressure 18 of the melting point. Okay. 19 This is something that I -- actually 20 my thermodynamics course, if you come to class 21 in two weeks we will cover. 22 Q. I will never understand it. 23 A. Well, the students don't all understand it 24 either, but obviously Dr. Anderson doesn't 74 1 either. He has the sign reversed. It is not 2 just he has got the magnitude off by a couple of 3 hundred miles down behind the center of the 4 earth. He has got the sign reversed. 5 Q. You are talking about his testimony page 165? 6 A. Right. He is throwing out scientific theories 7 as if he knew what he was talking about and it 8 is just not right. 9 Q. You don't have an asterisk or a Roman numeral 10 next to this but you written at the bottom of 11 the page, Page 180, any voids or porosity is 12 unacceptable? Is there a reason why you wrote 13 that down. 14 A. That's what he said. I think that is a near 15 quote. In fact, I wrote it down. I could have 16 put an asterisk by it. But it is something that 17 I'm not going to need to remember. It is so 18 absurd that to me as a welding engineer to 19 suggest that a solder joint or a braze is not 20 going to have inherently some porosity when you 21 are using a binary alloy or a ternary alloy. 22 Anything other than a pure metal, like copper, 23 which is what they have in MAPP probe, you will 24 always get some porosity. 75 1 In fact, Dr. Tom Siewart, Tom's 2 doctoral thesis at the University of Wisconsin 3 in 1972 was basically doing brazing and 4 soldering experiments on sky lab up in outer 5 space to try to see if you could get rid of 6 porosity when you didn't have gravity around. 7 Tom is now a scientist at National Bureau of 8 Standards or NIST in Bolder, Colorado and sort 9 of a friend of mine, which is why I happen to 10 know about his thesis, although I have read the 11 paper in the welding journal. But anyone who 12 knows anything about brazing or soldering, knows 13 that you expect porosity in the joint. In fact, 14 the one standard I can point you to, which is 15 the Copper Development Association has a 16 recommended practice, which I think has been 17 adopted as an ASTM standard now. But I would 18 have to go back and check. Basically says that 19 an acceptable solder joint for soldering copper 20 like the pipes in your house, would have no more 21 than 30 percent volume porosity. 22 Q. Who propagated that standard? 23 A. The Copper Development Association. This is a 24 group in New York City who are basically the 76 1 people who sell copper around the world. These 2 people basically have to deal with people 3 complaining about leaking in copper pipes. They 4 develop a standard saying if the plumber 5 produces more than 30 percent voids, it is a bad 6 joint, but anything 30 percent or more is going 7 to be acceptable. 8 Q. 30 percent or less? 9 A. 30 percent or less, I'm sorry. I misspoke. 10 Thank you. Is going to be acceptable. 11 There is actually a scientific basis, 12 in my opinion, for that 30 percent figure in 13 terms of leakage. Okay. It has to do with 14 whether you consider the porosity to be somewhat 15 -- what a metallurgist would call equiaxed, 16 nearly circular in extent. If it is elongated 17 like rat holes or ant tunnels or something then 18 you could get leakage with less porosity. 19 Physically the way the solder flows and braze 20 alloy flows, you are going to get things that 21 are going to have pores that are not elongated 22 but they are going to have dimensions that are 23 somewhat similar in terms of length and breath. 24 There is a reason why you choose percent. It 77 1 has to do with the volume filling of pores in a 2 two dimensional surface. 3 Q. And you believe that that is now -- or maybe an 4 ASTM standard, a 30 percent porosity? 5 A. Certainly some of the -- Copper Development 6 Association came out with this standard in the 7 early '90s. Partly in response to ISO 9000 type 8 of regulations that were coming ahead in the 9 industry. I know that some of those standards 10 that the Copper Development Association started 11 to promulgate have been adopted by ASTM. 12 Whether the 30 percent figure is in the ASTM 13 specs. I would have to go back and check. 14 Q. And that 30 percent figure in terms of, I don't 15 want to misdescribe or mischaracterize it, am I 16 right in saying that up to 30 percent porosity 17 is acceptable? 18 A. Right. 19 Q. That has to do with copper soldering or brazing? 20 A. It has to do with -- it is written for copper 21 joints by the Copper Development Association. 22 But the principles of the void filling and fluid 23 flow of the liquid solder or braze alloy is 24 exactly the same physics. Whether it is copper, 78 1 nickel phosphorous on steel or whether it is 2 lead tin with copper or whether it's tin-antimony 3 with copper; the physics of filling those joints 4 is exactly the same. 5 Q. So then, is it fair to say you disagree with Mr. 6 Anderson's opinions regarding the existence of 7 voids in the cross sections that he inspected in 8 his report? 9 A. Well, Dr. Anderson certainly found some voids, 10 but I do disagree on whether all the things he 11 talks about are voids. Specifically are the 12 larger ones are in fact voids. You will find 13 that they tend to migrate towards the corners of 14 that thing. And the reason for that is that is 15 the thickest section. That is where physically 16 if you look at the pressure effects, the surface 17 tension pressure effects, driving the flow of 18 the liquid as it fills the joint, that's where 19 the voids are going to want to concentrate due 20 to the metalastic head, the pressure generated 21 by the flowing metal. Those are voids. They 22 are located in an innocuous area, at the corner. 23 In fact, just that horizontal section meets the 24 DOT 39 spec. The vertical section on the other 79 1 side, also by itself meets the DOT 39 spec. You 2 actually have two joints, two DOT 39 joints in 3 this connection even if you have the void there. 4 But, what I disagree with about this porosity is 5 the fine -- what he calls fine porosity at the 6 interface, which he says is a bad bond. 7 Q. Are you going to refer to Dr. Anderson's report? 8 A. Not his report. But his exhibit, Exhibit 6 to 9 his deposition. He has some of his -- he has 10 some photographs, some of which may be in his 11 report. But he has a more complete set in his 12 notes. 13 There is a set of photos that have a 14 cover sheet metallography, of brazing material 15 in MAPP gas cylinders. So you come through here 16 and you see the cylinders that he had cut up at 17 Annomet, we will see here on figure 2B, you will 18 see one of these voids right at the corner. You 19 will notice that is the thickest region of the 20 braze. What I'm saying that's where you expect 21 the pour to migrate to. Surface tension forces 22 will pull it to that location. The thinner 23 area, the braze alloy wants to fill the thinner 24 areas of this gap. 80 1 Actually, what you will see down here 2 is you will see another void at the bottom, you 3 will see a thinner area in between. What 4 happens is this didn't get enough braze alloy to 5 fill all the way to the bottom. It started up 6 here where the meniscus would be. And the braze 7 alloy sucked down but it didn't have enough 8 braze alloy to completely fill the gap. The 9 little dot of braze alloy couldn't do it. 10 In fact, you can see where you don't 11 get the meniscus how the meniscus has been 12 pulled in because of braze alloy is filling the 13 gap. I can show you others where we do have a 14 meniscus because you had a little larger dot of 15 braze alloy. Since the whole joint is 10 times 16 rather than four times the length of the 17 thickness, it is all okay. We have more than 18 enough leeway safety factor. 19 Then you get to the next picture which 20 shows a close up of that elliptical void in the 21 corner. Then you see right here on the edge, at 22 what we call the bond line or fusion line 23 between the braze alloy and the steel. 24 Q. Is that Dr. Anderson is calling the meniscus? 81 1 A. No. The meniscus was this region right over 2 here on figure 2B. That meniscus is right in 3 here. You see kind of what he calls the 4 undercut. The undercut meniscus. That is 5 irrelevant, because the strength of the braze 6 joint is this region of where the braze joint is 7 between this two pieces of steel. The meniscus 8 doesn't have any particular strength to this 9 joint. The lack of a meniscus or an undercut at 10 a meniscus is irrelevant to the strength of this 11 joint. What I see on the next photo, which has 12 a scale bar of 10,000ths of an inch in the 13 corner, you see the big void; that is a void. 14 Then you see some little black specs right along 15 the bond line. Dr. Anderson says that is 16 porosity. Well, I told Mr. Ergo yesterday that 17 I would like to get this sample and look at it 18 in the scanning electron microscope. Because I 19 will bet a substantial amount of money, or 20 whatever you want to call it, that these are not 21 voids but these are iron phosphites inclusions. 22 You will notice -- frankly, they can't all be 23 voids, because you will see some of these black 24 specs over here in the steel. We are now saying 82 1 that the steel is porous after it has been 2 rolled down into a sheet. I don't think so. 3 Q. It is iron phosphites? 4 A. I believe they are going to be iron phosphites. 5 I believe. That's what you expect to form in at 6 the interface here with the iron. They are not 7 voids. The reason I know they are not porosity 8 voids is they are not all circular. Surface 9 tension requires that the voids try to form a 10 circle if they can. You will see the circular 11 nature of this elliptical void. The surface 12 tension forces that this dimension exceeds the 13 gravity forces by a factor of about a hundred to 14 10,000. Okay. So surface tension wins this 15 battle. That's why Tom Seiward didn't find an 16 affect when he got rid of gravity in sky lab up 17 in space. I actually told NASA not to spend the 18 money. But I was just a student at the time. 19 But nonetheless, the surface tension 20 forces, you can get all of this out of the Bond 21 number, B-O-N-D, it is not actually named for 22 bonding but it is a dimensional number used by 23 chemical engineers to look at surface tension 24 forces to gravity forces. I use it. If you 83 1 want to take my welding course, I use it to 2 describe how the fluid flows and the brazing in 3 soldering operations. 4 When you get to these dimensions, 5 things are going to want to form spheres if they 6 can. This is elliptical because it is trapped 7 in a channel between two pieces of steel. So it 8 can' t be a perfect sphere. But wherever it can 9 it tries to turn into a sphere which is on the 10 ends in the braze, liquid braze alloy. 11 But you look at these guys, these 12 small ones. They have a greater tendency to 13 form a sphere than the big one does. The big 14 one may be -- the surface tension forces may be 15 a hundred times greater than gravity. But in 16 the little ones, the surface tension forces are 17 a million times greater than gravity. So they 18 are definitely spheres if they are voids. I 19 look at these right now and I can tell you they 20 are not spheres. Look at that, that has got 21 corners on it. Those are inclusions. They are 22 intermetallic inclusions from the braze alloy 23 reacting with the steel, but they are not voids. 24 Q. When you say they are inclusions, is that a 84 1 biproduct? 2 A. It is not an open space. It is actually 3 different metallic space. You can think of it 4 as plums in a plum pudding, but these are pretty 5 small plums. They are a different material. 6 You can think of it as raisins in rasin bread. 7 It is a solid material surrounded by other 8 material. It is not a gab. 9 Q. How are those inclusions created, are they in 10 the brazing material, are they created by the 11 interaction of the metals? 12 A. They are created by the reaction of the 13 phosphorus, I believe, in the braze alloy with 14 the iron in the steel. It actually shows, not a 15 weak bond but an excellent metallurgical, which 16 we also know because of the contact angle here 17 of the voids itself. It is just how you 18 interpret the metallography. I think the way to 19 kind of referee this is to put it in scanning 20 electron microscope where a void will look like 21 a void and an inclusion will look like a piece 22 of solid material. You either have a hole there 23 in the microscope or you don't. I will bet we 24 don't have a hole. 85 1 MR. EPSTEIN: I'm going to finish up 2 with your notes with regard to Dr. Anderson's 3 testimony. Why don't we take a quick break. 4 (Break in the proceedings.) 5 BY MR. EPSTEIN (CONT'D): 6 Q. You wrote down in your notes, you didn't write 7 anything next to it in terms of an asterisk or a 8 Roman numeral, but you wrote page 171, he would 9 like a more uniform hardness in this critical 10 brazed region, what is the significance of that 11 testimony to you? 12 A. I didn't really put an asterisk or anything else 13 beside it. It is a technical point he made. I 14 will say that I am not 100 percent convinced 15 that Anomet did the micro hardness measurement 16 properly. When the braze joint is so much 17 stronger than the steel and the failure occurs 18 in the sheet steel. I don't care about the 19 hardness. If the failure didn't occur in the 20 braze, then it doesn't really matter what the 21 hardness is. 22 Q. Then it is irrelevant to you? 23 A. It is irrelevant. 24 Q. Okay. You wrote down on one line, page 139 with 86 1 an asterisk next to this. "If I make the 2 brazing material really strong then what I'm 3 going to do is fail the cylinder." Do you know 4 why you wrote an asterisk next to that? 5 A. You actually already asked me that. I said I 6 agree 100 percent with that statement. 7 Q. It looks like I have gone through your notes? 8 A. On Anderson. 9 Q. On Anderson. 10 A. I was following you, you went through them all, 11 you got them all. 12 Q. I see a couple of other asterisks in your notes 13 with regard to other deposition testimony. I 14 will ask you about those then we will to take a 15 break. With regard to the deposition of Sonia 16 Dunn-Ruiz, which is Mr. Shalaby's wife. Taken 17 October 16, 2007, you wrote Page 24, "Her 18 husband held the cylinder up to his ear to check 19 for leaks that might -- it seems to cut off with 20 an asterisk next to it. Can you tell me the 21 significance of that? 22 A. That night. 23 Q. What is the significance of -- why did you write 24 an asterisk next to that? 87 1 A. Just that it demonstrates that is good practice. 2 It is actually not something I usually do. It 3 is something that I think is recommended in the 4 BernzOmatic torch warnings, it is actually 5 something I don't usually do. I know I'd smell 6 it better than I'd hear it. It may just be me. 7 Q. In your notes regarding the deposition testimony 8 of Warren Ratliff, you wrote down page 31, torch 9 may not have been fully engaged on cylinder and 10 you wrote an asterisk next to that? 11 A. That is a pretty important point potentially. I 12 mean, I don't know if it's true or not. That's 13 what he's testifying to and it is something that 14 I would want to remember. This is early on. 15 These are some of the things that I got 16 relatively early in this whole investigation. 17 It is just collecting points. It seems to be a 18 significant point if the thing is not tight. 19 Q. Did that bit of Mr. Ratliff's testimony play in 20 significantly to any of the opinions that you 21 expressed in your report? 22 A. Be careful when say play in significantly. That 23 can show up different ways on the record. 24 Depending in significantly is one word or two. 88 1 Q. He is sharp. Factor in significantly? 2 A. Exactly. I am trying to being humorous. 3 Q. It almost worked. 4 A. In any case, what was the point again? 5 Q. Why don't you read. 6 A. This might not have fully engaged on the 7 cylinder. Not when I read the rest of the 8 things and started hearing about two and a half 9 inch long splits and bent five and ten degrees. 10 I don't think this was a small leak through a 11 incompletely screwed down cylinder. This was a 12 sudden release of gas. I think this was pointed 13 out yesterday. I think I said it in my own 14 report later on page five of my June 23rd 15 report. I said the fire ball is consistent with 16 a sudden release of pressurized liquid through 17 the burst split in the shell. This was not a 18 small leak. 19 Q. In conjunction with the deposition testimony of 20 Andrea Shalaby, you have a double asterisk next 21 to the testimony on page 45 of the transcript, 22 which reads he doesn't know what warning would 23 have prevented this injury? 24 A. Yeah, I mean if warnings are going to be an 89 1 issue here, now you got him saying he doesn't 2 know what warning would have told him not to do 3 whatever he was doing. I'm not one to be 4 talking about warnings here. That was early on. 5 Some cases I do describe whether warnings would 6 have prevented an incident had they been needed. 7 But I have learned that there are plenty of 8 other people who talk about warnings in this 9 case. So I will defer to all the other warnings 10 experts. 11 Q. Here are you notes back. The last thing I have 12 from your file. It is on stationery if you 13 will, US Department of Transportation Pipeline 14 and Hazardous Material Safety Administration. 15 It is dated October 31, 2005. At the very top 16 it says DOT-SP6686. It is signed for a Robert A 17 McGuire, associate administrator for the 18 hazardous materials safety division, I guess. 19 In any event, I believe you told me off the 20 record this is a document that was given to you 21 by Mr. Gentry at Worthington? 22 A. Yes. I requested it actually. 23 Q. Can you describe what this document is? 24 A. It basically gives Worthington or the folks in 90 1 Chilton, because I don't know exactly the dates 2 of when ownership transferred or whatever, it 3 gives them permission to not use the pure copper 4 braze that is in specification 39. The 5 rationale for that is if you going to the 6 welding literature you will find that copper and 7 acetylene will form an explosive compound. So 8 you should never use pure copper in contact with 9 acetylene. Now, methylacetylene is a form of 10 acetylene and so someone decided that in order 11 to prevent this hazard of forming a copper 12 acetylide that maybe they should use something 13 that had less than 60 percent copper. Which is 14 what you do for acetylene fittings in pure 15 acetylene. 16 So the request for MAPP gas was -- the 17 request was made from the DOT by the people in 18 Chilton to deviate from the specification 39 for 19 MAPP gas. Not for propane or for other things 20 but for MAPP gas. 21 Q. Can we make a copy of this append this as an 22 exhibit to the deposition? 23 A. Sure. 24 Q. I will give you your original back. 91 1 MR. EPSTEIN: This will be exhibit No. 2 3. 3 MR. ERGO: Mark, if I'm not mistaken, 4 I think I produced that to you. 5 MR. EPSTEIN: I won't you didn't. I 6 don't recall having seen it. 7 MR. ERGO: I won't guarantee you did. 8 BY MR. EPSTEIN: 9 Q. Can you tell me, did this play -- did this have 10 any significance to the opinions that you 11 expressed in either your report or supplemental 12 report? 13 A. Only in the sense that I think on page three of 14 my original report, I said that the cylinder 15 design meets the requirement of CFR17865 with 16 the exception of the braze alloy. That is the 17 exception. It actually does meet the 18 requirements of the DOT, but not that specific 19 CFR spec. They have got a waiver on the CFR. 20 It is DOT approved, but the way I wrote that 21 sentence -- it wouldn't have been correct unless 22 I had the exception. I needed that document to 23 back up my statement. 24 MR. EPSTEIN: All right, why don't we 92 1 go ahead and take a break to grab a sandwich. 2 (Break in the proceedings.) 3 BY MR. EPSTEIN (CONT'D): 4 Q. Dr. Eagar, going back to the other three cases 5 in which you were retained by Worthington 6 Industries, the Sandoval case, the Glenn case 7 and the Barrett case. Take those briefly one at 8 a time, did you issue a report in connection 9 with the Barrett case? 10 A. No, not yet. 11 Q. Have you rendered any opinions in connection 12 with the Barrett case? 13 A. No. Mr. Ergo and I have discussed it. 14 Q. Do you at this point in time hold any opinions 15 about the cylinder at issue in that case? 16 A. I haven't actually seen the cylinder yet, so 17 really no. We have some opinions from their 18 expert stuff, I guess. It is not final reports 19 yet from anybody. 20 Q. The preliminary would you say? 21 A. There was some settlement meetings that they 22 had. PowerPoint presentation on theories and 23 stuff. 24 Q. How about the Glenn case, you have inspected the 93 1 cylinder at issue in that case, correct? 2 A. Yes. 3 Q. Had you prepared a report in connection with 4 that case? 5 A. I think I have. I would have to go back and 6 check. I think I have. 7 Q. As you sit here now -- 8 A. I take that back. I'm not sure I have seen 9 their report. I should have seen their report 10 first. I think have I drafted a report, how 11 about that. 12 Q. Have you reached any conclusions in connection 13 with the Glenn case regarding the cylinder? 14 A. It was a failure in the sheet metal. We have 15 pictures of the bend and we measured the bend. 16 To it is a bending failure, similar type of 17 abuses. What is probably the case in this case. 18 Q. Okay. So you basically anticipated by next 19 question. To make sure the record is clear, is 20 it your opinion that the failure of the cylinder 21 in the Glenn case was a result of abuse? 22 A. Yes. 23 Q. Can you tell me what is the evidence upon which 24 you base your opinion that the cylinder was the 94 1 subject of abuse in the Glenn case? 2 A. The melt deformed before the steel sheet metal 3 broke and there is a substantial bend angel to 4 the nozzle. It is just a sound bite; the basic 5 reason. And the fact that he testified that he 6 was using it as he said to tap something. To 7 tap it with that bend means he was using it as a 8 hammer. Probably not the appropriate use of a 9 torch and cylinder? 10 Q. When he testified that he used it to tap 11 something, was it with the torch attached? 12 A. Yes. And a cigar in his mouth. 13 Q. Do you remember what he testified he used it to 14 tap, what was the recipient to the tap? 15 A. He was tapping a copper angle that he was trying 16 to solder on, it was slipping off and he was 17 tapping it back on. 18 Q. Was the torch on as he was tapping. Was flame 19 coming out of the tip? 20 A. As I remember in his deposition it was on. I'm 21 not sure it clear from the early statements but 22 in his deposition I think it became clear it was 23 on. 24 Q. I realize the word tap can be a bit ambiguous in 95 1 terms of how much force is being applied. But 2 is it your opinion that Mr. Glenn was going 3 beyond tapping the object? 4 A. Well, if we assume that it was 12 foot pounds, 5 30 foot pounds, that is a pretty good size tap. 6 Q. To put 12 foot pounds of force on the torch, 7 would that be for akin to hammering? 8 A. It depends where you hit on the torch. Whether 9 you are using the black body of the torch. Or 10 whether you are using the tip of the tube. I 11 can't remember right now. I would have to go 12 back and look at his testimony of where he was 13 hitting to. To put 12 foot pounds on the black 14 body would be like hammering. If you put it on 15 the tip, it would be like a good tack hammer, a 16 good whack with a tack hammer. 17 Q. Was there any testimony in the Glenn case as to 18 how long of a period there was between, as you 19 use the term, tapping and the failure? 20 A. Well, there were only eye witnesses who saw the 21 big flame shooting out of the hole that he was 22 working out of the hole in the ground. It 23 essentially was simultaneously with the tapping 24 as far as we could to tell. I guess I would 96 1 say. Again, I wasn't there. There was no one 2 who was looking directly at him when it 3 occurred. 4 Q. Was he in a manhole? 5 A. He was actually working out on the side of the 6 street and in a little hole. They had dug up a 7 pipe and he was laying on the ground on his 8 stomach reaching down into the hole, so the 9 torch was down in the hole. 10 Q. Had the depositions of percipient witnesses been 11 taken in that case, not Mr. Glenn himself? 12 A. Yes, they have taken some depositions. 13 Q. Did you review those depositions? 14 A. Some of them. Maybe all of them. I don't 15 remember right now. 16 Q. To best of your recollection, did any of those 17 fact witnesses testify that they were watching 18 Mr. Glenn at the moment the incident occurred? 19 A. No. As I said one women heard the whoosh and 20 turned around. She was across the street or 21 down the street with her granddaughter or 22 something. The woman in the house heard 23 something and looked out and saw a big flame. 24 There were a couple of people who were attracted 97 1 to the loud noise. This was not a big ball of 2 the fire type of thing. It was more of kind of 3 shooting straight up in the air. At least 4 that's the way they described it. 5 Q. With the Sandoval case, did you examine the 6 cylinder in that case? 7 A. That is an interesting question. There were 8 five or six cylinders. We dropped down to three 9 at the end. It wasn't clear which cylinder it 10 was or what torch it was. There were certain 11 few inconsistencies about what equipment was 12 being used in chain of custody. 13 Q. This was some question of which cylinder was the 14 incident cylinder in that case? 15 A. Yes. I did examine a cylinder and a torch. 16 Q. Were you able to feel comfortable in your own 17 mind that the cylinder that you examined in that 18 case was the incident cylinder or was that still 19 a question after your examination? 20 A. Certainly, it was still a question. The 21 cylinders didn't have any breaches in them, 22 which is maybe why was Worthington was released 23 from that case. They didn't have any cylinders 24 that had holes in them or anyone describing a 98 1 hole in the cylinder in that case. 2 Q. There was no cylinder that had a hole in the 3 Sandoval case? 4 A. Not that anyone described. Not in the ones that 5 we had or that anyone describe as a whole. 6 There was certainly some missing cylinders in 7 that case. 8 MR. ERGO: There was never an 9 allegation that there was a breach in those 10 cylinders. 11 MR. EPSTEIN: Okay. Thank you. 12 Q. Other than this case, the Shalaby case, the 13 Barrett case, the Glenn, and Sandoval case, am I 14 correct that you haven't been retained by 15 Worthington in connection with any other 16 litigation involving an NRT cylinder? 17 A. That's correct. 18 Q. Referring back to your initial report in this 19 matter, Exhibit No. 1, if you look at section 20 number 4, in there you state, I'm paraphrasing, 21 that the steel was sound and that no 22 deficiencies are present. That there is no 23 evidence of a deficiency in the steel. 24 Can you explain what you relied on in 99 1 making that statement? 2 A. Well, I'd say that the very fact of forming this 3 thing from a flat sheet of the steel into a 4 cylinder means that you had to have stressed the 5 steel beyond its yield strength. And not only 6 stretched it beyond it, but formed it and 7 stretched it had made it go through substantial, 8 what we call, ductility. That forming process 9 provides the evidence of this ductility. So the 10 very nature of taking a flat sheet of steel and 11 forming a cylinder that is going to be brazed is 12 a proof test that the steel is of good quality. 13 It would never make it to the brazing furnace if 14 it broke during the forming process. 15 Q. Is it possible to stretch flat sheet steel 16 beyond the yield point and have a resultant 17 weakness in the steel from that stretching that 18 isn't necessarily a break in a or breach in the 19 tensile strength of the steel? 20 A. Let's just say it would be statistically 21 improbable. If you had something that was that 22 close to the failure of the steel, when you are 23 stamping out thousands of these things from the 24 same coil of steel, you would have a reject rate 100 1 that would be so high that everyone would take 2 all those cylinders and reject them just because 3 your reject rate was so high. In the forming of 4 these thing, the reject rate would be very, very 5 low. In terms of splits or anything that occurs 6 during forming. If you had something that was 7 that close to failure, then would you expect 8 reject rates 5, 10, 20 percent range and that 9 would raise a big red flag that that steel was 10 not of sufficient quality. 11 Q. Is it possible to have steel within in a coil, 12 sheet steel that has varying qualities to it? 13 The beginning of the coil is one quality and 14 middle is a different quality and the inner part 15 of the coil is a different quality still? 16 A. Not in the way we make steel today. Even in the 17 olden days, I mean like the 1920s and stuff. 18 You really didn't have that kind of problem with 19 the steel. The steel is made in 300 ton lots 20 today. The olden days maybe it was made in 150 21 ton lots. It goes into a 20 ton ingot and then 22 gets rolled down and stretched out by a factor 23 of a hundred or a thousand. And going through 24 all of that, you deform it and you homogenize 101 1 it. You put it through furnaces to homogenize 2 it. So you get very uniform consistency within 3 a coil. It may have variability from 4 coil-to-soil. But within a coil you have very 5 good consistency. 6 Q. In making the statement that, I'm paraphrasing 7 in item number four on your evidence, there is 8 no evidence of deficiency in the steel. You 9 didn't conduct any tests personally upon which 10 you base that statement, correct? 11 A. I'm talking generally in the steel here when you 12 are just the generic. We don't have the actual 13 cylinder here. If that's what you are asking. 14 Q. I understand we don't have the actual cylinder. 15 You didn't conduct of any exemplar test of 16 anything to reach that particular conclusion? 17 A. No. But I do know that up here it says DQSK. 18 Q. Yes. 19 A. The steel mill has already gone through 20 substantial quality control procedures to make 21 sure that this is the very best of formability 22 of the steel they produced. They have already 23 done tests. They are shipping this to the 24 automotive companies or the appliance companies 102 1 like General Electric in Louisville. People buy 2 million of tons of steel from them. They have 3 to maintain a very high level quality. The DQSK 4 is based on the highest quality of formability 5 of the steel that you can purchase. 6 Q. Do you know what the tensile properties of that 7 steel are? 8 A. The tensile properties are required by the DOT 9 39 to be like 30,000 pounds by square inch. 10 However, in a modern steel mill, that is very to 11 meet, because we have what we call fine-grain 12 practice. The steel is made typically in a 13 continuous caster. These are all changes that 14 occurred in the last 30 years. If we were 15 talking about the 1960s or before, you might 16 have some variability, but the modern steel 17 making practices for the last 30 years have 18 actually produced a very, very high quality 19 drawing steel. 20 Q. Referring to item or section number five in your 21 report, which is at the bottom page three, goes 22 on halfway to down page four. Can you describe 23 the particular requirements for the braze joint 24 that connects the main valve to the cylinder? 103 1 A. Well, if you read the DOT 39 where it has to be 2 four times -- the length of the braze has that 3 to be four times the thickness of the steel. As 4 I pointed earlier during the morning, that is 5 not an uncommon type of requirement when you are 6 talking about making braze joints or solder 7 joints in sheet metal. You want to have a 8 substantial overlap, a four-times overlap, if 9 you will, compared to the thickness. That 10 requirement comes out of CFR 178 that the DOT 39 11 that we talked about. 12 Now, it turns out that what we have 13 here is actually in this design something that 14 is not a factor four overlap, but it is more 15 like a factor ten or 12 overlap. You have 16 plenty of safety factor. In addition, the four 17 times overlap is actually based on a pure copper 18 braze where the copper is substantially weaker 19 than the steel. We actually have a braze alloy 20 that is comparable in strength to the steel. 21 Therefore that further increases the safety 22 factor. So that is where -- those are some of 23 types of things that take this from a safety 24 factor of two or three on strength to a safety 104 1 factor of 20 on strength that we talked about 2 this morning. 3 Q. What tests are conducted at the Chilton, 4 Wisconsin facility that are designed to test the 5 braze joint? 6 A. Well, there is a periodic hydrostatic burst 7 test. I think it is specified in the DOT 39 of 8 how many every thousand or whatever cylinders. 9 As I remember discussing with Mr. Gentry, the 10 practice is do them as about twice as often as 11 DOT requires. If you do one a day, you do two a 12 day, it might be 10 a day. As I remember, Mr. 13 Gentry told me it was about twice as often the 14 DOT requirement. That is where you test one to 15 failure. You pressurize it. That is called a 16 hydrostatic burst test. Then you have a leak 17 tests on every single cylinder that is sold. 18 They go through a water bath, a hot 19 water bath. In the hot water bath, if they 20 detect any leak, this is after they have been 21 filled, the braze has been made. They have been 22 filled with MAPP gas. If they detect any leak, 23 they will take them out of service. Obviously, 24 you are not going to sell a leaking tank, 105 1 because it is not necessary that it is so 2 hazardous but when the customer finally gets it 3 at Home Depot it is going to be empty if it is 4 leaking. Almost any leaks that you are going to 5 find is going to empty that thing within weeks 6 or a month. In shipment you might leak out a 7 little bit of gas, a slow leak is actually not 8 that hazardous. 9 Q. Why is a slow leak not hazardous? 10 A. You would have something ignited. If it 11 ignited, it would only be a small little flame, 12 smaller than a match head, probably. 13 First of all, where is the source of 14 ignition in transportation. Even if it got 15 ignited a small wind would blow it out. Even if 16 did ignite, it wouldn't be a great big fire 17 ball. It is going to be a small leak by 18 definition. If it is a big leak, you are going 19 to pick it up at the plant an never ship it. 20 So even if a small leak existed, you 21 can't have too many of them or you are going to 22 have customers complaining, I just a bought a 23 tank and it doesn't weigh, two pounds, it weighs 24 one pound, which is just the weight of the steel 106 1 and nothing comes out. I don't think that 2 occurs very often. I never actually heard of it 3 occurring, would you have to ask the folks at 4 Chilton if it has ever occurred. It is not 5 really a serious hazard, even if you have a 6 small leak. I'm not sure where we were going 7 with a small leak beyond that. 8 I'm sort of rambling. 9 Q. Are you aware of any tests that are conducting 10 at the Chilton facility to test the tensile 11 strength of the brazed joint from the exterior? 12 A. That would be the periodic hydrostatic burst 13 test. And in fact, it's my understanding and it 14 would be expectation, I think I did talk to Mr. 15 Gentry about this that you do get a break in the 16 cylinder wall and not up at the nozzle when you 17 pressurize these things. Each one of them when 18 it goes through that hot water tank is 19 pressurized and I think Mr. Gentry testified 20 that it is 319 PSI which is a 50 percent over 21 pressure above -- more than a 50 percent 22 overpressure about what you would expect to ever 23 see in service. Even if this thing gets hot, 24 like 140 degrees in Palm Beach or something, 107 1 sitting out in the sun in some desert, you would 2 never get close to the type of pressures that 3 you would see in the testing that they do. 4 Q. How large did you say the safety factor is? 5 A. 319 over 188 as I remember. But the point is, 6 the hot water tank is substantially hotter than 7 the desert and it is the temperature that 8 generates the pressure. Very few deserts are 9 1880 degrees. At least not the deserts I go to. 10 Q. Referring to Section 5 of your report, what is 11 the release rate of the pressure release safety 12 valve? 13 A. Release rate? You mean release pressure? 14 Q. Yes. 15 A. It is something above 319. As I remember. I 16 think it is one of Mr. Gentry's depositions or 17 somewhere. Something below 400 but above 319 as 18 I remember. I seem to remember 375 but I 19 wouldn't want to testify to that under oath. I 20 guess I am. But no, I'm not. I'm telling you 21 that I'm not sure of it right now. 22 Q. The last paragraph in section 5 of your report, 23 what are the strength and quality requirements 24 that you refer to? 108 1 A. Last paragraph in number five. Let's see. 2 Right here? 3 Q. Yes. It reads, "Thus, the manufacturing 4 geometry and the quality control test..." 5 A. Okay. The strength requirement is the 30,000 6 pounds per square inch of DOT 39. In fact, by 7 doing the periodic hydrostatic burst test, you 8 prove that it's well above that. I can't 9 remember, I think I calculated once if you were 10 at the minimum DOT strength you might burst at 11 700 PSI. I calculated it once. And it is well 12 below the 980 or whatever that they actually 13 test that. Which means the steel is stronger 14 than it. The quality requirements basically it 15 is the quality of do very many of them leak. If 16 you have a lot of them leaking, you are going to 17 stop your manufacturing process. Every single 18 one of them has to be tested before it goes out 19 the door. If you have got a lot of leakers, you 20 are going to go in there and fix the process 21 because you are just losing money. Okay. You 22 can't afford to just manufacture things that 23 aren't going to be sold. 24 Q. You state, "It is improbable that a 109 1 manufacturing deficiency is significant enough 2 to result in a failure of the cylinder and 3 service will ship from the manufacturing 4 facility. 5 Can you state the basis for that 6 statement? 7 A. Sure. I have to demonstrate it for you. I have 8 to do my History Channel demonstration. You 9 will find me on the History Channel if you watch 10 the story of the Titanic. I did this. 11 Basically, if you have a flaw in the material, 12 it will weaken the material. What I typically 13 do for my students, and I did on the History 14 Channel, if I take a piece of paper I can pull 15 on the edge of that piece of paper with several 16 pounds of force. If I introduce a defect or a 17 flaw, and I try to tear that, it takes ounces 18 not pounds. So that little defect weakens that 19 piece of paper by a factor of 10, let's say. 20 Well, if we are going weaken this 21 cylinder sufficient to cause a failure in 22 service, we are going to have a significant size 23 defect. That defect is going to be something 24 that will probably be caught, if you had one, by 110 1 the leak test or the burst test things. It 2 would have to be a substantial size defect. All 3 these quality control tests, basically show that 4 you have your process in control such that the 5 probability, what we call in manufacturing the 6 -- they actually talk about different levels of 7 Sigma, which is a standard deviation of a 8 probability curve. The Motorola kind of coined 9 the term in quality of Six Sigma. Maybe it 10 wasn't Motorola, but they won the Malcolm 11 Baldridge quality award from the Department of 12 Commerce in 1989. That's where I first heard 13 about Six Sigma when I was the Motorola liaison 14 professor. 15 But it's generally attributed to 16 Motorola. Six Sigma means you can have only 17 three flaws out of a million pieces. Three cell 18 phones out a million that might not be work 19 properly. That's what you are shooting for is a 20 6 Sigma quality level; it is what people shoot 21 for in the industry. With these types of 22 requirements and things and doing 100 percent 23 inspection, your shipping is probably well below 24 or well above Six Sigma, it might be seven or 111 1 eight Sigma. Now, that is partly because you 2 test every single one. As I remember, Dr. 3 Anderson had a -- might be one of his very last 4 exhibits in Exhibit 6. Somewhere when I was 5 going through this, I saw a failure rate of how 6 many leakers they found. 7 The fact is, you are going to catch 8 those leakers you when you go through the test. 9 The number of the shift is down approaching 10 zero. I don't know if it is actually zero. The 11 ones that actually could, after they have gone 12 through all of those tests and go out into 13 service, you might have something that might 14 leak and empty, but you don't have anything that 15 is going to have be big enough flaw to cause a 16 catatropic thing that is going to cause a 17 fireball that injured Mr. Shalaby. The 18 probability of that is less than the number of 19 gains of sand in the sea. It is extremely low. 20 I have never heard of it. 21 Q. A short time ago you testified that if you were 22 to have a cylinder with a leak that was able to 23 ship from the facility, and if you were to 24 introduce an ignition source, it would probably 112 1 result only in a -- I don't want to 2 mischaracterize your testimony, but a small 3 flame or a flicker of a flame. 4 A. The size flame that is going to have enough heat 5 to sustain itself, is typically going to have to 6 be about three-eighths or a half an inch high. 7 If you have a smaller flame than that, just the 8 heat sucked out of it by the surrounding steel 9 where the leak is liable to quench the flame and 10 put the flame out. You actually have to have a 11 reasonable size leak in order to get a flame if 12 it got ignited that would self-sustain itself? 13 Q. Is it possible in that instance where you have a 14 leak for it -- the leaking gas to act in the 15 manner of a fuse where the heat or the flame 16 might actually work its way into the cylinder, 17 ignite what is inside? 18 A. You are actually talking about something that is 19 flashback. If you look the combustion 20 literature or the welding literature. In fact, 21 I'm giving the plumber lecture at the American 22 Welding Society and about 25 percent of my talk 23 is flashback and how we teach flashback. I 24 could have brought my lecture which I finished a 113 1 few weeks about. But flashback requires that 2 you have a source of oxygen as well as fuel in 3 order to get the flashback. There is no oxygen 4 in that cylinder. By the very nature of which 5 they filled them, there is no oxygen. There is 6 certainly not enough oxygen to sustain 7 combustion and get a flashback or a fuse effect 8 like you are talking about. Even if you did get 9 heat back in there, there is no oxygen to burn 10 in there. The vapor space is virtually pure 11 hydrocarbon. Even if you had a little bit of 12 oxygen. As I remember, the upper flammability 13 for methylacetylene is something like 15 14 percent. You would have to have 85 percent air 15 in that thing. 16 It is just thermodynamically 17 impossible at room temperature to not have more 18 than 15 percent methylacetylene vapor in that 19 area space. I shouldn't call it air space. In 20 the voids space above the liquid inside that 21 cylinder. I think we can prove 22 thermodynamically it would impossible to get an 23 explosive mixture inside that tank. 24 Q. Absent what a breach in the cylinder wall? 114 1 A. If you have breach in the cylinder wall, now you 2 have all the methylacetylene in the air. That's 3 where the explosion is. It's the fire ball 4 outside of the cylinder. When people explode 5 cylinders or have explosions breaches of 6 cylinder. Take the two tests in Mount Shasta. 7 You didn't have a flashback back into the 8 cylinder. It was the liquid coming out that had 9 the fireball out surrounding the cylinder. 10 There was no bulging of those cylinder after 11 Mount Shasta, because there wasn't a burning 12 inside the cylinder. Inside the cylinder you 13 are too rich in your fuel air mixture to get an 14 explosion inside. 15 You can prove that from the relatively 16 high vapor pressure of methylacetylene at room 17 temperature. It is going to be much more than 18 15 percent. You would have to introduce liquid 19 air with the liquid MAPP in order to get a 20 flammable mixture inside. That's not have how 21 they filled them. There is no liquid air 22 around. 23 Q. In your report, you reference the depositions 24 that you have reviewed in this case. On page 115 1 one, is that the complete list of the 2 depositions that you reviewed? 3 A. Except for Robert Anderson and Mrs. Vredenburgh. 4 Q. In reviewing the deposition testimony of Mr. 5 Ratliff, the campground manager. Did you review 6 his testimony about interviewing a guest at a 7 neighboring campsite who, according to Mr. 8 Ratliff's testimony said that he or she, 9 observed Mr. Shalaby banging a cylinder against 10 something hard? 11 A. I have in my notes on page 20, Mr. Ratliff said, 12 "The woman at the next camp site said Shalaby 13 was banging the torch on the fire ring. And on 14 Page 21 he said that the husband of the woman 15 confirmed the banging. That's my notes without 16 going back to the actual pages. 17 Q. How important is that testimony to the opinions 18 that you've expressed in your June 23rd report? 19 A. I don't know if I'd use the word important. It 20 is consistent with my conclusion that the only 21 way I can get the breach that was testified to 22 by not Mr. Ratliff, whoever the other ranger was 23 Stephens, the only way you can get that breach 24 and the bend that was testified to is by abuse 116 1 of the cylinder. You have to go beyond the 2 normal use level as Dr. Anderson called it. 3 Now, what is the type of abuse. This 4 is consistent with the type of abuse. I will 5 let some other person determine whether they 6 believe Mr. Ratliff or Mr. Stephens. Their 7 testimony would come in and other people can 8 decide whether they believe that testimony. I 9 can tell you that it takes a substantial banging 10 type of force, like I have described it before, 11 somewhat dramatically, like using it as a hammer 12 to get this type of damage. You don't get this 13 by just waving it around the air or even tapping 14 it against the table? 15 Q. What about after -- assume for the moment that 16 cylinder is purchased at some point in time and 17 the corrosion process is permitted to begin in 18 the braze joint area somehow, we will leave 19 those factors unknown for now. In approximately 20 a year after the date of purchase, the cylinder 21 still has contents and it is attached to a TS 22 4000 torch and picked up by the user, is it 23 possible in that year worth of time if -- is it 24 possible in that year worth of time that the 117 1 corrosion process could weaken the braze joint 2 to the point that simply it picking up with 3 torch attached and moving it around with the 4 torch attached could result in a breach? 5 A. Well, you haven't really quantified the amount 6 of corrosion. Let me put it this way: Mr. 7 Shalaby testified that he took the torch off 8 when he stored it in the trailer each time he 9 used it, as I remember, after he used it. If 10 that's the case and you've ever actually seen 11 one of these things that was in a human 12 environment and got rusty, for example, in the 13 Sandoval case, I took the little plastic cap off 14 of one these things and the unpainted portion of 15 the cylinder was just brown with rust. It was 16 thick with rust. The painted portion of that 17 same cylinder was not rusty at all. This little 18 corrosion area that Dr. Anderson talks about is 19 painted. 20 If Mr. Shalaby had taken this thing 21 out of his trailer to screw the torch on and it 22 had been corroded with rust, he would have been 23 screwing it onto something that there would have 24 been so much rust he probably couldn't have 118 1 screwed it together. The thing that is going to 2 rust first is the unpainted threads. You would 3 never get enough weakening of the wall, the 4 painted wall near the braze joint to corrode 5 through to weaken it before you would get such 6 rusty threads that no one in their right mind 7 would screw the torch onto the cylinder. 8 Q. Is the testimony of the two, I'm go going to 9 call them paramedics, the EMT personnel, Mr. 10 Russo and Price, who testified about about Mr. 11 Shalaby saying that he kicked the cylinder into 12 the fire. Does that -- can you describe to me 13 the role that that testimony plays in the 14 opinions that you formed in your June 23rd 15 report, if any? 16 A. Let me just say that whatever Price said in his 17 depo, I didn't think it was significant enough 18 to make a note about it. With regard to Mr. 19 Russo, I do have on page 22 he says, "Shalaby 20 said he kicked the cylinder." Okay. That's my 21 one note among the two depositions. 22 Frankly, I have been asked whether 23 kicking it into the fire ring or banging it 24 against the fire ring is more probable in terms 119 1 of what these people said. If I had to pick 2 one, I will take banging hammering, because it 3 is consistent with the angle of the bend that 4 the fire -- park rangers talked about. It is 5 consistent with the two-and-a-half inch split. 6 If it is kicked into the hire ring and the fire 7 ring is hot, then it would be consistent with 8 hearing a hiss, which both Mrs. Shalaby and Mr. 9 Shalaby said they heard a hiss of leaking gas 10 before the explosion. If I had this thing 11 sitting on a hot fire, I could build up pressure 12 and I could blow off the pressure relief valve. 13 It should blow off. 14 Q. Would that create a hissing sound? 15 A. That could create a very loud hissing sound 16 compared to the type split of you get that was 17 described by the park rangers. So I really 18 can't tell you as a scientist whether banging 19 against the fire ring or kicking into the fire 20 and heating up in the fire and the pressure 21 relief letting go, I don't have enough evidence 22 to come in and say scientifically this is 23 consistent. Either one of them is abusive of 24 the cylinder. You shouldn't throw it into a 120 1 fire, you also shouldn't be banging it like a 2 hammer. Okay. Now, it is also possible the 3 split, the park rangers saw, occurred 4 afterwards. I could have kicked it into the 5 fire ring and never had a split. But I could 6 blow that pressure cylinder off, that pressure 7 relief off. I could hear a hiss. If you have 8 ever seen one of these things ignite, when you 9 have got that stuff streaming out of a thing 10 like that, I have got a video that the acetylene 11 guys put together. The compressed gas 12 association. I sometimes show it to my class. 13 It is very dramatic, you see four, five, ten 14 foot flames; depending on the size of the 15 pressure relief. 16 Q. Coming out of the pressure relief valve. 17 A. The problem I have with that is Mr. Shalaby 18 described being engulfed in a flame in the 19 middle of the fireball. Shooting out of a 20 pressure relief valve is more like being hit 21 with a flame thrower as opposed to being 22 engulfed in a fire ball. If it was coming 23 straight at you, maybe. 24 Then I have to factor in where his 121 1 burns were. If you at his hands, his hands are 2 fairly severely burned. That is consistent with 3 banging it like a hammer. It is not as 4 consistent, Well, it got kicked into the fire 5 ring? Well, how did he get his hands so burnt, 6 was he putting them in front of the flame 7 thrower. All I'm really saying is all of this 8 testimonial evidence, whether it is Russo, 9 Price, Ratliff, Stephens, Mr. Shalaby, Mrs. 10 Shalaby, I can't give you a, "Oh, here is the 11 scenario that fits all of this." Okay. I could 12 rely on certain pieces of it. I could say, "Oh, 13 yeah, it got kick in the fire ring, had a 14 pressure relief let go, you had a flame thrower 15 type of shootout. 16 I could also say, Well, hit it like a 17 hammer against the fire ring. Do you want me to 18 take a choice? It's not a scientific choice. 19 If I had to take a choice, I shouldn't take a 20 choice if it's not a scientific choice. But if 21 you want, I think banging against the fire ring 22 is more likely, because it is simpler. I don't 23 have to then have the fracture by his throwing 24 it afterwards. But, you know, I wasn't there. 122 1 It is not for me to determine. I can say as a 2 scientist all of the evidence points to abuse. 3 But exactly what caused the abuse, is not for me 4 to say. 5 Q. As between two the scenarios that have been 6 suggested or proposed or testified to, that is, 7 being Mr. Shalaby either kicking or tossing the 8 cylinder into the fire versus banging it or 9 striking it against a concrete fire ring, do you 10 find, in your mind, the more probable scenario 11 to be Mr. Shalaby banging the cylinder; is that 12 what your testimony? 13 A. Using what is called Ockham's Razor, which is 14 the simplest explanation is the best. That is 15 Ockham's Razor. 1292 AD. In any case Ockham's 16 Razor basically favors banging against the fire 17 ring, it explains the split, it explains the 18 bend. 19 But, you know, for me to start parsing 20 out whether Mr. Stephens, Mr. Ratliff, Mr. 21 Price, Mrs. Shalaby, Mr. Shalaby, Mr. Russo. 22 Whose testimony is most believable, I can't pick 23 between heating up in the fire ring or banging 24 against -- banging against the fire ring or 123 1 heating up in the pit. 2 What I will tell you is if I had the 3 cylinder I could tell you. I could distinguish 4 the two if I had the physical evidence, I 5 guarantee you I could. I will give you better 6 than 90 percent probability that I could tell 7 you the difference between those two if I 8 actually had the physical evidence. 9 There is another third scenario. That 10 is waving it around or the pulling the trigger. 11 I can eliminate those based on the forces 12 involved. I don't get 30 foot pounds or 12 foot 13 pounds or anything else by swinging this around 14 or pulling the trigger. I'm off by a factor of 15 a hundred in my scientific forces. 16 Q. Okay. You have led to another line of 17 questions, I was going to raise, which is: 18 Getting back to the issue of corrosion, I know I 19 already asked you about that. You testified 20 about how the dreaded area would be the first 21 area to the exhibit signs of corrosion? 22 A. Right. 23 Q. If the brazed joint -- if corrosion were to 24 occur in the area of the brazed joint, is it 124 1 possible that the corrosion could weaken the 2 strength of the joint such as you wouldn't 3 require 12 to 30 foot pounds, I might be off on 4 my range of force in order to create a breach? 5 A. If you have enough corrosion, you can I 6 perforate the thing and start a leak. Obviously 7 somewhere before you start a leak that it has 8 gotten so thin that is going to be weakened. 9 However, as I said, the unpainted area 10 corrodes first and the corrosion on that 11 unpainted area is going have to be something on 12 the order of creating a rust thickness that is 13 on the order of three-eighths of an inch thick. 14 Are you have to rust away 30,000ths of steel or 15 28,000ths, whatever the wall thickness of this 16 thing is. It is about 28 to 30,0000th up in 17 this region. There is typically a eight to 10 18 fold increase of volume of rust versus the steel 19 it came from. 20 So the rust coating on those threads, 21 would gum up those threads so you could never 22 put the torch on, okay. You would have to be 23 cleaning the threads off with a wire brush. 24 Every time you put the torch on, and you would 125 1 have do that four or five times. 2 The other thing I'll tell you. As I 3 understand it, this cylinder was only about a 4 year and a half old. The typical corrosion 5 rate, for unpainted steel out in the rain, is 6 4,000ths per year. 7 Q. 4,00ths? 8 A. 4,000ths of an inch. Which means it would take 9 seven years, if you left this unpainted in the 10 rain, exposed to the elements, it would take 11 seven years to perforate this thing. Which 12 means it would take five or six to get a leak or 13 to weaken it substantially. It is not that old. 14 From that point, the corrosion theory 15 is scientifically impossible. Unless, you soak 16 it in salt water, you can't corrode it fast 17 enough. I go back to the what is the known 18 corrosion rate, atmospheric corrosion rate of 19 carbon steel. 20 If you want a reference, you can go to 21 ASM International puts out a book called 22 Corrosion Data. I'm sure you have a copies at 23 home. I have one at home. I have one at work. 24 I think I have two at work. There is also the 126 1 Metals Handbook, volume 11 or whatever is 2 corrosion in the 10th edition. That will have 3 the data. 4 Q. I believe earlier that you testified that you 5 cannot remember the name of professional society 6 that came up with a figure of 30 percent voids 7 being acceptable level? 8 A. The Copper Development Association. 9 Q. Do you accept that as a valid standard? 10 A. Yes. 11 Q. So then applying it to the case at hand, is it, 12 would you opine that it is acceptable in the one 13 pound NRT cylinder that are manufactured at the 14 Chilton, Wisconsin facility to have up to 30 15 percent voids in the brazing material? 16 A. I would think that is a reasonable standard. I 17 don't know that I've ever seen that standard or 18 that they have ever worked that standard out. 19 There is actually other reasons why that would 20 be a fair standard. It has to do with the fact 21 that the void space is going to congregate in 22 the corner which is the thickest area. It is 23 not going to be randomly distributed across that 24 L-shape joint. It going to be concentrated in 127 1 the corner of that L-shaped joint. 2 Q. In reviewing the photographs or included with 3 Dr. Anderson's report, is it your opinion that 4 those photographs or none of them depicted a 5 lack of fusion between the brazing material and 6 the parent material? 7 A. That's correct. They actually depicted very 8 good fusion. You can actually see the effects 9 of some of the erosion of the base material. 10 Plus, I would point out, that given the fact 11 that those little dots of solder started up on 12 the outside rim of that L-shaped joint and the 13 braze made its way all the way down around the L 14 to the bottom, if I didn't have good wetting, 15 W-E-T-T-I-N-G, if I didn't have good wetting of 16 that braze alloy, it would never get down all 17 the way to the bottom. We know it got down to 18 the bottom in every joint that anyone has ever 19 looked at to my knowledge. 20 Q. Generally speaking, in a generic sense, is a 21 lack of fusion between brazing material and 22 parent material acceptable? 23 A. No. Let me also say it would show up, I have 24 talked to Mr. Gentry about this, it would show 128 1 up as those four dots of braze alloy, would not 2 flow together and make a continuous circle on 3 the top? 4 Q. If there was a lack of fusion? 5 A. If there was a lack of fusion, those four dots 6 would remain as four separate dots or three dots 7 or the circle would only go halfway around, or 8 two-thirds of the way around. In fact, Mr. 9 Gentry explained to me, that is one of the 10 visual inspection techniques that people are 11 told to look for in Chilton, Wisconsin to make 12 sure that they see the braze alloy has flowed 13 all the way around, 360 degrees around the 14 clock. 15 I might add that's the same thing that 16 the US Navy came up with in the subsafe program 17 when the Thresher submarine went down in 1962 18 for their braze joints. So nuclear submarines 19 use the same inspection technique as Chilton 20 does or the people at Chilton do. 21 T-H-R-E-S-H-E-R up in Portsmouth, New Hampshire. 22 Q. Does aging of cylinder, absent corrosion, aging 23 and normal use, have any weakening factor on the 24 braze joint? 129 1 A. There is no substantial effect on the weakening 2 of the steel itself. Not in these types of 3 steel the DQSK. Which is what we call. The SK 4 stands for special killed. Today's terminology 5 that means alluminum killed. Which means there 6 is no strain aging effect in these fields. We 7 used to have problems with that in the 1930s. 8 But not in the 1960s. Not since well before the 9 first half of the 20th Century. There is no 10 mechanical property change or brittleness in the 11 steel. Plus we wouldn't have a bend if there 12 was a brittleness of the steel, an angular bend. 13 In addition, Dr. Anderson sort of said 14 that the joint degraded, I think is in my notes 15 somewhere. That as -- I remember him saying, 16 page 119. He thinks the torch degraded with 17 time. If that's the same type of thing we are 18 talking about. Other than corrosion, which I've 19 already discounted, it would take five or six 20 years for unpainted steel, this is painted 21 steel, there is no degradation mechanism of 22 steel that substantially weakens it over a one 23 period of time. If there were, you should buy a 24 new car very three months, because that same 130 1 sheet steel is going into your car. If the 2 steel degrades and becomes substantially weaker 3 by a factor five or ten or 20, than your car is 4 going to fall apart every three months. It is 5 the same steel. 6 Q. Assuming that corrosion is occurring in the 7 braze joint, corrosion would have -- aging would 8 have a factor on the strength of the steel of 9 some sort? 10 A. Aging just means corrosion goes deeper. It 11 takes five or six years, even out in the rain. 12 If you store it inside your trailer, it is not 13 in the rain rather than five or six years, it 14 might take 50 or 60 years. 15 Q. Do you agree that an older cylinder that has 16 been out in distribution for some period of time 17 may not be able to pass the same tests that it 18 was subjected to at the time it was newly 19 manufactured? 20 A. Only, if it has been stored in salt water. 21 Q. Only if it has been stored in salt water? 22 A. I would bet that my six-year old cylinder that 23 is sitting in my office now, would pass any test 24 that would be given to it, would pass any of the 131 1 DOT requirements today. If you emptied the MAPP 2 gas and pressurized it and things like that. 3 MR. ERGO: You could probably think of 4 others, other than salt water. 5 A. That is true. But is the one that has been 6 suggested in this case. 7 Q. Turning to Section Eight of your report. Maybe 8 I wrote it down wrong. It must be six. 9 A. Six is fairly brief. 10 Q. Section six? 11 A. Sorry. I was page six, I'm sorry. 12 Q. Let's turn to your supplemental report dated 13 July 28th in Section Eight of that report. 14 Which is fairly brief. What is it about the 15 British study that you were supplied with that 16 suggests to you that the braze joint between the 17 valve housing unit and the cylinder is not the 18 weak area of the cylinder? 19 A. Because even though they had flames impinging on 20 these cylinders full of gas, they had all kinds 21 of other things that they did. They never had a 22 failure at the nozzle connection. So they did 23 all kinds of abuse, very severe abuse. If you 24 want to talk about other abuse Mr. Ergo just 132 1 mentioned, there is a hole series of tests. 2 Every abuse they could think of they subjected 3 these things to. They never got a failure of 4 the brazed joint in the nozzle. Which is just 5 confirmation that that is a very strong braze 6 joint. 7 Q. Turning back to the first page of your 8 supplemental report. In the bottom paragraph, 9 you are critical of Dr. Anderson on a number of 10 things, no stress analysis or testing of the 11 strength of a braze joint was performed. Do you 12 believe that a stress analysis or a testing of 13 the strength of the braze joint was conducted at 14 the Chilton, Wisconsin facility? 15 A. I don't know if they did a stress analysis in 16 the computer or not. But testing, every time 17 they do a hydrostatic burst test, which is 18 several times a day they are doing a test of the 19 braze joint. 20 Q. Is that a stress test of the braze joint? 21 A. It's a stress test of the entire cylinder. It 22 doesn't fail the braze joint. The braze joint 23 is not the weakest area. He claims it is the 24 weak element in the assembly. The hydrostatic 133 1 burst test, which they do twice a day up in 2 Chilton shows that is not true. All the tests 3 done by the British shows that is not true. 4 Q. In your opinion should the Chilton manufacturing 5 facility conduct any additional tests to test 6 the tensile strength of the braze joint from 7 tests that apply force from the exterior? 8 A. Additional tests. I don't know what test they 9 might have done. They may have inherited this 10 design from 50 years ago for all I know. There 11 may not have been any history of problems that 12 tells them, Oh, you have got to go do a why 13 stress test or a stress calculation on it. 14 You don't usually go out and test 15 things that have had decades of successful 16 service just to spend money. You usually do 17 stress analyses when you are designing something 18 for the first time. When this was designed 19 there were no computers. This is before the age 20 of computers. You would have to have something 21 that says there is a problem with the braze 22 joint to go do a stress analysis on that braze 23 joint. I have never seen a failure of a braze 24 joint other than the one they overheated up at 134 1 Mount Shasta. 2 Q. Aside from the failed cylinders you've seen in 3 connection with the four litigation matters in 4 which you have been retained, have you been 5 provided any data by Worthington Cylinders about 6 failure rates, not testing done at the facility, 7 but failure of cinders that have been shipped to 8 market? 9 A. I don't know if I want to say it is provided by 10 Worthington. But both in this case and -- 11 certainly in this case. People have listed like 12 there have been nine failures or 24 failures or 13 20 failures. I have heard numbers of nine 14 failures out of 24 million. 20 failures out of 15 24 million. I don't know what they mean by 16 constituting a failure. I think there are nine 17 things that Mr. Anderson went through of other 18 cases where supposedly there was a problem. 19 But I think they included Sandoval in 20 there. Where there wasn't even a breach of the 21 cylinder. So you really can't say, We have nine 22 failures. In Sandoval there was no failure of 23 the cylinders. You really have to know more 24 detail about these things in order to say that 135 1 these were quote, unquote, failures, where I'm 2 going to call a failure here, I'm going to 3 define it here for the purposes of the Shalaby 4 case as a catastrophic sudden release of a large 5 amount of gas. 6 Q. What I meant, have you been provided with any 7 statistics from Worthington or anybody else 8 about whether -- for example, do you know if 9 there have been any failures of NRT cylinders 10 produced at that the facility resulting in leaks 11 of gas from five to ten years ago; do you have 12 any data on that? 13 A. I don't have anything from more than five years 14 ago or more than 10 years ago. Let's say. I 15 have heard about nine since 2000 or something. 16 Those the kind of the things that I think Mr. 17 Anderson has testified to, things in his exhibit 18 6. 19 Then Mr. Ergo told me that apparently 20 there are 20 releases or something out of 24 21 million cylinders produced since September of 22 2004. 23 Q. 20 releases at the facility? 24 A. I don't know. I don't have anything other than 136 1 I have heard the number of 20 out of 24 million. 2 Again, I don't know what it means by a failure. 3 Was the failure a big fire ball or was the 4 failure just the leak that was -- that got out 5 into the world. 6 Q. In my case I'm referring to a leak, just a leak, 7 that didn't necessarily result in a fireball? 8 A. Okay. 9 Q. Okay. Putting aside any statistics you may have 10 or may not with regard to tests conducted in the 11 manufacturing process where cylinder rejected. 12 I'm talking do you have any statistics, were you 13 provided any regarding known failures of NRT 14 cylinders produced at the Chilton, Wisconsin 15 facility that have gone out into market besides 16 the information that is referenced in Dr. 17 Anderson's report? 18 A. No. 19 MR. ERGO: I just want to note for the 20 record, I don't know where the number 20 came 21 from. I don't recall that being something I 22 told you. 23 Q. Did Mr. Ergo or anybody else tell you whether 24 they have data regarding known failures of NRT 137 1 cylinders that were shipped to market outside of 2 information regarding matters referenced in Dr. 3 Anderson's report? 4 A. We did discuss that sometime last fall. It was 5 my understanding there were at that time, six, 6 seven, eight, nine something like that. 7 Q. Total? 8 A. Total. 9 Q. That was information provided to you verbally? 10 A. Yes. By that, I mean, basically I think that 11 was probably cases. Again, Sandoval was a case. 12 There was no breach of the cylinder. You can be 13 sued without having a release of gas from the 14 cylinder. 15 Q. Right. Inverse of that, you can also have a 16 release of gas from the cylinder that doesn't 17 result in a lawsuit? 18 A. Yes. 19 Q. I'm asking what information you have or don't 20 have on known cases where cylinders have failed 21 resulting in a gas leak. It doesn't have to be 22 catastrophic in the sense there was a fire ball 23 or anybody got injured, where cylinders were 24 shipped to market and it was a known leak? 138 1 A. It is my understanding that you can basically 2 count them on your hands, you don't have to take 3 off your shoes. 4 Q. This information that was provided to you by Mr. 5 Ergo? 6 A. Yes. And reading Dr. Anderson's deposition. 7 The numbers are consistent. 8 Q. Did you ask Mr. Ergo or anybody else at 9 Worthington if they have any data with regard to 10 known failures of cylinders that were shipped to 11 market that resulted in gas leaks? 12 A. I don't remember asking that question. 13 MR. EPSTEIN: Let's go off the record 14 for a moment. 15 MR. EPSTEIN (CONT'D): 16 Q. Let's go back on. Dr. Eagar, getting back to 17 the issue of porosity in the brazing material, 18 would you agree that a void that exists in the 19 brazing material -- 20 (Interruptions in the proceedings.) 21 Q. A void is a bubble, a space where there is no 22 brazing material? 23 A. Yes. 24 Q. In that space where there is no brazing material 139 1 within a void, where one exits, there is no 2 fusion or anything to hold the valve housing 3 unit in place within that space? 4 A. Within the void, that is correct. Void does not 5 add strength. 6 Q. A void to one extent or another takes away 7 strength; is that correct? 8 A. Well, that gets to be an interesting question 9 whether it detracts from strength of the 10 assembly. It depends on how you define your 11 assembly. If your assembly includes the sheet 12 metal, then who cares whether on 30 times as 13 strong as the sheet metal or 20 times as strong. 14 A chain is a strong as its weakest link and the 15 weakest link is the sheet metal in the cylinder, 16 then you have a chip, degraded the strength at 17 all by having a void. 18 Q. By what you are saying has a matter of degree. 19 I'm not talking about whether it is acceptable, 20 unacceptable. Where a void exists -- let me say 21 it differently. When there is one or more voids 22 in the brazing material as opposed to no voids, 23 you have some reduction of strength in that 24 area; is that correct? 140 1 A. There is no strength in the area of the void. 2 Whether that decreases the strength of the whole 3 assembly, gets to be what is the weakest link in 4 the assembly. If the braze alloy is not the 5 weakest link in the assembly it makes no 6 difference. If the brazed alloy is the weakest 7 link, then it would degrade the strength. 8 Q. In terms of the strength of the bond between the 9 -- I have heard it referred to different things, 10 engineers yesterday called it bushing, I'm 11 talking about the -- you called it something 12 else earlier today. 13 A. The valve. 14 MR. ERGO: Worthington calls it center 15 valve housing. 16 Q. I want to make sure we're not talking about 17 apples and oranges. Where there exists a void 18 in the brazing in the joint between the central 19 valve housing unit and the cylinder wall, that 20 adds to some degree, whether you can measure it 21 or not, of a weakening of the -- again, the 22 joint. The bond between the valve housing unit 23 and the cylinder wall? 24 A. A joint with a void is not as strong as a joint 141 1 without a void. Whether that degrades the 2 strength of the cylinder depends on whether the 3 joint is the weak link. 4 Q. At some point, would there be a -- strike that. 5 If the number of voids within the 6 brazed joint area got to some point, would you 7 have concern regarding the strength and the 8 integrity of the unit. In other words, beyond 9 30 percent or some other number? 10 A. Since we have essentially a double joint here, 11 in terms of DOT requirement, the horizontal and 12 the vertical both meet the requirements. You 13 could probably in this particular joint and 14 given the fact that it is 20 times stronger by 15 my estimate than it needs to be. You could 16 probably have a 70 percent void and still have a 17 braze that is stronger than the sheet metal. 18 The concern that I would have is if you don't 19 get a good joint going around 360 degrees 20 because that is going to result in a leak. That 21 is going to be picked up because you do 100 22 percent leak testing. 23 Q. In your opinion, is it possible that the 24 incident at issue in the Shalaby, in this case, 142 1 could have been caused by an inadequate brazed 2 joint in the area between the valve housing unit 3 and the cylinder? 4 A. No. 5 Q. Do you feel comfortable ruling that out? 6 A. I see no way a flaw in the braze joint 7 contributes to abuse. And abuse is what caused 8 this thing to fail. 9 Q. Do you feel comfortable to opine that the 10 failure of the cylinder at issue in this case 11 could not have occurred in the absence of abuse? 12 A. Based on the ranger's testimony of what they saw 13 on the cylinder afterwards, yes. 14 Q. If it turned out the rangers gave testimony on 15 the issue or were inaccurate in their 16 description of what they observed on the 17 cylinder, would you be able to stand by your 18 opinion based on the other evidence in the case? 19 A. There has to be some significant breach of the 20 wall to release the fireball as suddenly and 21 catastrophically as it did. If they tell me 22 that no, they didn't see a two and a half inch 23 long crack but they saw a circular hole or 24 something, I would want to know more details 143 1 about that circular hole. I know there had to 2 be a substantial breach in order to cause the 3 fire ball. This was not a small leak. 4 Q. We have in our office, back in California, a -- 5 one of a few old MAPP gas cylinders that Mr. 6 Shalaby had purchased years ago. Until recently 7 it had an audible leak, you could put it up to 8 your ear and you could hear -- describe it as a 9 bubbling sound. I'm not sure how else to 10 describe. Mr. Ergo heard it. Is it your 11 opinion that a leak of that nature would not be 12 result in the incident which occurred in this 13 case? 14 A. It wouldn't have a catastrophic fireball that 15 would burn him on his hands arms and legs. A 16 small leak is not going to be able to do that 17 type of damage unless he wants to take a few 18 hours and move it up and down. 19 Q. Use of a cylinder that has a leak in the nature 20 that I've described to the best of my ability, 21 couldn't result in a fire ball, absent abuse, in 22 the type that injured Mr. Shalaby? 23 A. That's correct. 24 Q. I believe you testified that the matters in 144 1 which you've worked with Western -- were 2 retained by Western Manufacturing involved one 3 pound NRT cylinders that were used for propane; 4 is that correct? 5 A. Yes. 6 Q. It was your estimate there were three or four 7 such cases? 8 A. I remember one in Canada, one in Mississippi, 9 Missouri. I don't remember if there two others. 10 I remember there was a fourth but I wasn't the 11 primary expert. Okay. They wanted to use a 12 younger person. 13 Q. Did any of those cases involve releases of 14 propane gas where the user was burned? 15 A. Yes. 16 Q. Did all those cases involve such incidents? 17 A. Certainly the one and Canada and the one in 18 Missouri did. I think the others did as well. 19 I don't remember the details of the other two as 20 well. 21 Q. Did you render opinions in those cases as to the 22 cause of the leaks? 23 A. Yes. 24 Q. Was it your opinion that involved abuse by the 145 1 users? 2 A. Well, if you consider throwing it to a concrete 3 floor as hard you can abuse, yes. 4 Q. One of them said? 5 A. They both kind of said, I knocked it off the 6 counter and it fell to the floor. Except we 7 actually had at cylinders and demonstrated by 8 tests on exemplar cylinders, that the only way 9 you could get the bending, the breach was with 10 an impact on the order of 15 foot pounds or 20 11 foot pounds. Not all that different than what 12 has been measured by Dr. Anderson and Dr. Carr. 13 If you then take their description of 14 the accident, it was on the shelf, I went to 15 reach for it, and I knocked it on the floor. If 16 it just fell to the floor, it had to drop 45 17 feet, not four feet, in order to do that type of 18 impact damage. 19 On other hand, if they that grabbed it 20 and they actually grabbed something that was hot 21 and they threw it out of their hand, I was 22 explained throwing something at 30 or 45 miles 23 an hour or whatever I calculated was not a Roger 24 Clemens' fast ball, they are not professional 146 1 baseball players. But if someone is trying to 2 get something hot out of their hand, they could 3 probably throw something at 40 miles an hour. 4 Which was consistent with the impact damage. 5 That is basically the way approached both of 6 those. They basically claim the thing fell. I 7 think one time six feet and one time four feet. 8 I showed it has to be 30 or 40 feet. The speed 9 at impact due to the damage, the physical 10 damage, had to have be something equivalent, I 11 don't remember if it was 30 or 45 miles an hour. 12 Something. It was something that you could 13 imagine any normal person throwing something to 14 the ground to get rid of something hot in their 15 hand, it could hit the concrete with that 16 velocity? 17 Q. Getting back for a moment to the Shalaby case. 18 Is it possible that the breach which caused the 19 accident could have been caused by the user 20 having the torch and cylinder in his hand and 21 putting it down too hard on the table. 22 A. If he puts it down like hitting a hammer or a 23 nail through a two-by-four, yes. 24 Q. Do you have a number in terns of PSIs or some 147 1 other type of measurement, how much force it 2 would take in terms of putting a cylinder down 3 on down a surface? 4 A. It is not a force, it is an impact which is 5 about 30 foot pounds according to Dr. Anderson. 6 Q. Do you agree with that? 7 A. The 15 to 30 foot pound range is consistent with 8 the types of numbers that I remember from the 9 tests that I did on the NRT cylinders in the 10 past. 11 Q. You in your mind cannot imagine, 15 to 30 foot 12 pounds of force placed on the cylinder by 13 somebody who is using the tool in a hurry and 14 puts it down quickly in the midst of 15 multi-tasking say? 16 A. No. They have to be -- it is not something that 17 you do with your wrist. It is not something 18 that you do with your wrist and your elbow. 19 This is something that you have to do with your 20 shoulder on. Okay. Let me just say that. In 21 order to get 15 foot pounds, you don't do that 22 with your wrist. Not even Arnold Schwarzenegger 23 does it with his wrist. Maybe Arnold could do 24 it with his elbow. I can't do it with my elbow. 148 1 I have to use my shoulder into it. I have to 2 have leverage of the length from my shoulder to 3 my hand. 4 Q. Dr. Eagar, if the cylinder had been kicked into 5 the fire ring with the glowing ambers that Mr. 6 Shalaby's wife testified -- the state of the 7 fire that they testified that it was in at the 8 time the accident occurred. Do you have an 9 estimate it would take for the heat to go above 10 to a point to cause a gas release? 11 A. I haven't tried to estimate that. 12 Q. What temperature does it require, would it 13 require for the -- strike that. 14 What temperature is required to cause 15 the pressure to build to a point that the 16 release valve would have to engage? 17 A. As I said earlier, I don't remember exactly what 18 the pressure is the relief valve goes at. Given 19 the fact that they put it in 180 degree water in 20 order to do the test of every cylinder, and you 21 don't want the pressure release valve to go at 22 180 degrees. It has to be something above 180 23 degrees. I don't think it is 250 degrees. But 24 it is probably in the 200, 220 range. I haven't 149 1 -- I'm giving you an estimate. This is a very 2 rough estimate. I told you the basis for my 3 estimate. 4 MR. ERGO: I think we are going to 5 have to do a fact check. My memory is, I have 6 been told, that the water bath is 130 degrees. 7 A. Okay 130 degrees. If that is the case, maybe it 8 is 180 degrees. But that is the basis. It has 9 to be something more than the temperature of 10 water bath. Don't want to trip the pressure 11 release in the water bath. 12 Q. I understand. I guess it is clear, you 13 conducted no experiments on this issue? 14 A. I haven't really looked into that. I suppose I 15 could, but I haven't. 16 Q. Given the facts as you know them from the 17 various evidence that has been presented to you, 18 and given your knowledge as an engineer in the 19 field in which you specialize, does it make 20 sense to you that Mr. Shalaby could have been 21 injured by having kicked or thrown the cylinder 22 into the fire? 23 A. I think we discussed that. I would expect that 24 if you throw it into the fire and it is there 150 1 long enough as you have pointed out, that you 2 could have the pressure relief let go, and I 3 have pictures, I have videotapes in my office 4 showing what happens when a pressure release 5 lets go on an acetylene tank, for example. I 6 have done it. Not on an acetylene tank but it 7 was furniture polish. A little spray can of 8 furniture polish that fell off. Someone put it 9 on their space heater and went to fix dinner and 10 it fell off and burned their two-year-old child. 11 With a flame-thrower basically. We actually did 12 the test, it is pretty dramatic to see like 13 hairspray or in this case furniture polish, you 14 will get a six or eight foot flame shooting out 15 of that thing. But it is all directed in one 16 direction. It is coming out of a nozzle so that 17 is a little different that the fire ball. 18 Q. If that were to occur where the pressure release 19 valve activated the flame shoots through it, 20 would that result in the type of cracking that 21 Mr. Ratliff and Stephens testified that they 22 observed? 23 A. No, it would not. That's why you then have to 24 have them -- to get the cracking they observed 151 1 you now have to say maybe that is when he threw 2 it. Which, you know, I don't disagree with Dr. 3 Anderson when he says, Yes, throwing it. If he 4 throws it far enough and it hits the right way, 5 you might crack it. 6 Q. It would take a series of factors to come in to 7 play in order to create the crack under that 8 scenario; is that correct? 9 A. You wouldn't get this type of crack at this 10 location. You definitely wouldn't get it by 11 putting it in the fire. You get the pressure 12 relief valve to go first. If the pressure 13 relief valve, for some reason, didn't let go, 14 you'd get a split on the side. Just like they 15 did in the British test where they defeated the 16 pressure relief or whatever. You'd get the 17 highest stresses, the internal pressure 18 stresses, as we've discussed a couple of times 19 today, are on the side of the cylinder not at 20 the top. 21 Q. You wouldn't get by putting the cylinder into 22 the fire, however it got there, you wouldn't get 23 a fireball of the nature that Mr. Shalaby 24 described? 152 1 A. You would get a directed flame like a flame 2 thrower. If you were standing right in front of 3 that, you might think it as a fire ball. 4 Because you are not exactly standing there 5 staring at it. You are trying to hide your face 6 and everything else. It is not the type of 7 fireball that we saw in the video at Mount 8 Shasta. 9 Q. It wouldn't be the type of fireball that Mrs. 10 Dunn-Ruiz described when she turned around and 11 saw her husband on fire? 12 A. Exactly. It wouldn't be like that. In fact, 13 when it comes out of that pressure relief, there 14 tends to be a roar. It not like a jet engine. 15 And it is not a dig cat, it is a small cat, 16 roaring at you. It is not a lion. 17 Q. A cub? 18 A. A cub or something. It is noisy. Because it is 19 coming out of a nozzle. A fire ball is not 20 coming out of a nozzle, it is just surrounding 21 everywhere. It doesn't make loud roaring 22 noises. But something that comes shooting out 23 of a nozzle makes roaring noises. 24 Q. Given all the factor that we've just discussed 153 1 in the last few minutes, as you sit here now 2 does it make sense to you the scenario that Andy 3 Shalaby kicked the cylinder into the fire and 4 received the injuries that he did and the 5 cylinder received the damage that it did? 6 A. I can't rule it out, because they did say they 7 heard hissing. It is easier to hear hissing as 8 the predecessor to a flaming roar. But, I've 9 already told you, if I had to pick one, I will 10 take using it as a hammer on the fire ring as 11 opposed to kicking it into the fire ring pit. 12 I'm not going to tell you 13 scientifically it is one or the other, because I 14 just don't have enough information. 15 MR. EPSTEIN: Let's go off. 16 (Discussion off the record.) 17 MR. EPSTEIN: Let's go back on. Dr. 18 Eagar, thank you I have no further questions. 19 THE WITNESS: Thank you. 20 MR. CARRUTH: No questions. 21 (Time noted 2:28 p.m.) 22 23 24 154 1 COMMONWEALTH OF MASSACHUSETTS. 2 COUNTY OF SUFFOLK, SS. 3 I, Lori J. Atkinson, Professional Court Reporter and Notary Public duly and 4 qualified in and for the State of Massachusetts do hereby certify there came before me the 5 deponent herein, namely, THOMAS EAGAR, Sc.D., who was by me duly sworn to testify to the truth 6 and nothing but the truth concerning the matters in this cause. 7 I further certify that the foregoing transcript is a true and correct transcript of 8 my original stenographic notes. I further certify that I am neither an 9 attorney or counsel for, nor related to or employed by any of the parties to the action in 10 which this deposition is taken; and furthermore, that I am not a relative or employee of any 11 attorney or counsel employed by the parties hereto or financially interested in the action. 12 IN WITNESS WHEREOF, I have hereunto 13 set my hand and affixed my Notarial Seal this 17th day of September, 2008. 14 15 LORI J. ATKINSON NOTARY PUBLIC 16 My commission expires: 17 December 20, 2013 18 ***PLEASE NOTE*** THE FOREGOING CERTIFICATION OF THIS TRANSCRIPT 19 DOES NOT APPLY TO ANY REPRODUCTION AND/OR DISTRIBUTION OF THE SAME BY ANY MEANS UNLESS 20 UNDER THE DIRECT CONTROL AND/OR SUPERVISION OF THE CERTIFYING COURT REPORTER. 21 22 23 24 155