THE COMPRESSOR CASE.

Gisborne Times, Volume XLVIII, Issue 4601, 3 July 1917, Page 6

THE COMPRESSOR CASE.

ALLEGED DEFECTIVE

VALVE.

•J. J. NIVEN, LTD. v. P.B. FARMERS’ MEAT CO.. LTD. THE CASE FOR THE DEFENDANTS. The case for the defence in the case J. Niven, Ltd. v. P.B. Farmers’ Meat Co. was continued at the Supreme 'Court yesterday morning, when Mr Myers commenced to call liis evidence. The first witness was Alexander Stuart Mitchell, engineer, and member of the Mechanical Institute, said that for the past eleven vears he had been engaged exclusively on designing and supervising the erection of refrigerating machinery. In addition to his New Zealand work he had engagements in Australia and South America. During the last 16 years he had made a special study of compressors and valves and was familiar with the English and American types. He had on several ocoasions inspected the damaged compressor at Waipaoa and would put down the accident: to faulty design and construction. By construction he meant workmanship. His Honor: I take it you mean the workman’s error.

Witness: Not only that. He went on to say that he believed the design was defective. He considered the stop which limited the travel of the valve was faulty and the clearance was too small. ” By clearance it was understood the difference between the radio of the circumference of the mushroom head and the interior of the cage. The insufficient clearance would result in increased pressure in the cylinder beyond what it should be. It would have the effect of lifting the mushroom head from the seat further than was necessary and proper. The valve would come back with a rebound on the seat. Under normal conditions the stop should not be brought into use at all. He had nejvor experienced before the valve striking the stop as had apparently been done in this case. Bursting of a valve would be an extreme condition. The result of the striking the stop would be to “fatigue” the metal and would ultimately fracture it. A stronger spring would not overcome the difficulty and Avould simply produce a greater rebound on the seat, ultimately producing the same results as he had indicated. He did not think it possible to give the seqiience of operations resulting in the accident —it would be all speculation. Apparently the stop was put there for receiving impacts from the valve. His .Honor: Is it not there in all valves ?

CONFLICT OF EXPERT OPINION

Witness: Not in that way'. He went on to say that he knew all the well known types of valves. He knew of none with such small clearance as that in question. The relation between the cub'c content of the cylinder and the clearance was yery undefined. The 13 x3O stroke Linde compressor at Nelson Bros, was not of standard design, although made by Linde. Each Linde compressor was fitted with the same size valve without regard to the length of the stroke. That showed that the suggested relation to which reference had been made did not exist. He had not fathomed some of Mr Hunter* s figures: they did not seem to have anv bearing on the matter. That witness’s figures were an absurdity in themselves. He had seen Mr A. C. Michell’s figures in regard to getting gas velocity which he also regarded as valueless. As tho gas became further compressed the velocity became consequently reduced. His’Honor: You have got beyond me. Mr Myers: He means that the calculation should not be taken after compression. Witness went on to say that Mr Mitchell had not adopted the proper method of ascertaining the velocity or gas escape from the cylinder. He made it 173 feet per second as against 97 by Niven’s. The recognised formula was in this case: 324 divided by 4.47 and by 60 and then multiplied by 144. For steam 8000 ft was a safe velocity. In gas compressors makers went on the basis of 6000. He had worked out the relation in other machines as between clearance area and piston area. That was,. in his opinion, a reasonable comparison to make. In the Vilter it was just over 6 per cent, Hall’s was 7 per cent, Linde (at Waipaoa) 5.2 per cent, Haslem 8.6 per cent. Niven’s at Waipaoa worked out at 1.45 per cent. Mr Myers: Does that help to prove insufficiency of clearance? Witness: Certainly. QUESTION OF CLEARANCE.

As regards Mr Chadwick’s illustration there was no analogy as between .inn working conditions. The SpLfamnSfem to that witness had referred was put there foi repair work, etc. When tna. gas was allowd to pass through that particular half-inch pipe Alie machine had to be run very slowly. Those conditions required frequent stoppage to enable the machine to cool down. As to Air Walker’s statement that the gas could be put through a half-inch pipe, why was so much trouble taken to give so much extra delivery space in the machine itself? The valve opening was the most restricted portion through which the gas is required to pass, in Niven’s case more gas pressure became necessary as there was . rstricted clearance in the one in question, the most restricted area in the Niven valve produced was at the top of the mushroom head, when the valve was in course of lifting. Where Niven s had most restricted part it should be the least restricted. Niven s _lia_d made a. reversal in one of the principles of valve design as exhibited by Hall, Haslem, Vilter, York. Linde (British). Witness was shown a small new Linde valve. The most restricted portion of the valve in the half-ton Linde valve produced was not in the same place as in the Niven valve. In the Linde and Haslem valves produced the most restricted portion was at the valve seat. In the big Linde valve, produced, the same condition as he had described in the other two valves also existed. Valves from the Hercules and Stern machines were also produced, and in these witness said the same condition applied. • . Mr Myers produced for the jury s inspection drawings of valves snowing the clearance in the ordinary valves and the Niven valve.

WAS THE MISHAP L'NUSTJAL ?

After examining the drawings witness said he did not expect the Niven valve to work regularly and smoothly, but he expected the other would. The area of the clearance should increase as the valve lifted, but it did not do so in'the Niven valve. > This was shown by the diagram produced. Witness produced a working drawing of J. and A. Hall’s valve, in which the same principle was displayed. Referring to .velocities, witness did not consider a velocity of 173 feet pen second 'either 'safe or efficient. Witnss had known quite a number of . compressors and he knew of no acci- • dent caused by the mushroom, head of . a discharge valve dropping into a cylinder in other compressors. He had known of springs being in compressors for eight'or nine, years. With the valves as they were at the P.B. worKS before the breakdown witness did not think the machine could be worked at its full capacity. This was due to insufficient clearance. Had the valves been alright between 130 and 140 revolutions would bo required to produce the required amount of ice. Indicator card 26 was taken at ice making conditions rather , than fieezing . conditions. It was not asking too much from - the. machine to- take the ..

card. No. 15 card indicated excess pressure in the cylinder. An excess of 121bs was shown, and this 71bs in excess of which he would have ex-, pected. In card 26, there was an excess of 261bs, that was to say 211bs beyond normal. He should say that this indicated the gas was having a difficult job to get the cylinder. Speculation on the matter of time m which the valve head and other parts were reduced was idle as 'it was impossible to say. It was impossible to sav when the marks on the piston were made. They were quite different to the marks in the cylinder caused by the mushroom head. Assuming the marks on the piston were done at the same time the condition which brought about the marks did not contribute to the accident. He would hardly expect a rhythmic sound from the Niven machine. The number of valves could not all at once. .

To His Hpnor: The valves on, the. Linde did not seat at once, and he would not expect rhythmic sounds from machines where valves did not seat simultaneously. To Mr Myers: The Linde machine had one delivery valve at each end. The machine was a low revolution one and the design of the valve, he thought, would have to be revised for high revolution conditions.

EFFECT OF MACHINIST’S ERROR.

To Mr Skerrett: Witness was first a mechanical engineer and had served his time with a Sydney firm. After that he studied for the profession of architect, but he had never followed the profession purely. Witness had had no workshop experience or experience -in operation of machinery since his apprenticeship. He had never been in charge of refrigerating machinery, he had never designed refrigerating machinery and he had not designed a valve. Niven’s were recognised as first-class experts in the making of refrigerating machinery. That was until they began to start making valves of 'this design. Witness heard of the accident shortly after it occurred. He saw the type of this valve about,four,months ago. Mr Skerrett: Do you invite the jury to say that your knowledge in regard to Valves and compressors is superior to Niven’s experts? Witness: Yes.

Mr Skerrett r Although you have never worked a machine or designed a valve?

Do you say whether the machinist’s error in these valves is or is not negligible ? Witness: It is not negligible. Then if 'the machinist’s error had not existed would it have provided sufficient clearance ? Yes.

Then the very slightest increase of clearance would have been sufficient? I do not agree altogether. Have you not said it was vour opinion if the machinist had bored out the cage a little further the clearance would be sufficient?

Yes. That is what you refer to the defect in construction ?

The faulty nature of the stop which limited the travel of the valve. Was this a defect in construction or design? In design first of all. What other defects ?_ The materials of which the valves were made might be better. What? Do you say the quality was not good ? It was not suitable. The seating might have been made from steel instead of castiron.

WITNESS ON THE LOG BOOK

Continuing, witness said that insufficient clearance would at once begin from the start and he would expect the collar of the valve would strike upon the metal stop from the beginning. These conditions would be abnormal in the case of an ordinary compressor and valve. The first result of insufficient clearance •' ould be undue heating of gas, • and the delivery pipes would be heated more than would be expected in a proper running machine. He would not necessarily expect these indications to be shown on the return gauge. The engineer in charge would not notice an abnormal heating of the delivery pipe. Undoubtedly the lack of clearance was substantial, and he expected that the less clearance there were the more the heat. One hundred and twentyfive was approaching the maximum of the machine. - Mr Skerrett produced a table of workings taken from the log book. Witness thought the record showed a satisfactory operation of the machine. He could not say why here were no consequences from insufficient clearance recorded in five months, nor could he account for it. If there were any they ought to have been slightly reflected in the record of the running of the engine. 'lt was quite possible that the consequential results of the defect would not have eventuated within the five months. Mr Skerrett: Is that probable ?

Witness: It is unlikely. Asked to explain the marks on' the •jiston being caused by other than the impact of the piston with the mushroom head protruding through the valve opening, he said that some foreign matter might have got into the cylinder, hut he did not think so. The impact must have taken place a number of times. The force of the blow the piston would deliver upon the mushroom valve would be the force of the resistance of the valve. He thought that the spindle, might have been pushed back into the cage, lie did not think it was possible for the guide to hold the mushroom rigidly. If it were held rigidly he would expect heavy blows, but he did not think it was held rigidly. Beyond the marks •lie had seen he had noticed no other evidence that the piston rotated. There was evidence of rotation. it did not follow that piston and rod had rotated together as a result. <i the knocking on the mushroom head, -the piston and rod might have rotated L independently. It was quite feasible that vibration might have caused the rotation. ’■ ,

CAPACITY OF THE MACHINE.

Pressed by Mr Skerrett, witness said he thought that the theory put to him might be as feasible as his own. The piston might have rotated before the mushroom head had protruded through the valve opening. The piston rod screwed into l the cross!)ead 1 and the lock nut- was there to keep it secure. If the lock nut .did not loosI eh the piston would not rotate. The | impact would not on a light load make ■ a distinct noise, and he did not think i it would create a noise that a . competent ngineer would detect. If the piston hit the mushroom 'head white that was rigid} there would he a thud. Apparently the largest circumference on the mushroom head had been ! staved-up by. contact with the piston j head. A ‘series of light tappings would produce the same effect. (Alter i the trials there should be a period of : at least one month of maintenance so l as to have the machine under operaj tion. It was witness’s, practice to !inspect machinery before delivery, that was when the machine was manufactured in New Zealand. ■_ Mr Skerrett: Do you pledge your professional reputation that this com- : pressor, as delivered, was not capable of producing 36 tons of . ice per 24. j hours? - . „ .' :• Witness: For how.long? I Mr Skerrett: Say for five months ? i Witness: It was not capable. ! How long was it capable. - i I could not say. I. have never given ! the matter consideration. < How do .you -fix(five, months ? ! According to my opinion the maehi ine to have been capable of doing that ; would have required 130, to ,140 revoI * U you think this machine could ■. have been.run at from 130 to 140? i ■ ■ Apparently there was / some good 1 for not doing so. . , _ . Mr Skerrett repeated his question, - and witness replied that , it should iSSSII . iilfH

have been capable. He went on to say that there were other reasons, defective valves for instance. He did not think that even with running at 140'it would have produced the required amount. v< Had he, when inspecting the machine, ' found valves not up to requirements, and there,was another machine' available, and the stores were, empty, he would have stopped tin/ machine. .. .. Continuing, witness said it was highly improbable that the mtfsliroom head fell into the cylinder immediately after the fracture - some time must have elapsed; iirmay have been days or it may been hours. Assuming the spindle of the valve had been broken, he knew of no subsequent process by which the spring could be broken oi be caused to fail. He had it on authority that Professor Mclntyre m Ins book had set up. the standard that the ratio of clearance was 6 qier cent. Witness had found it so in practice. Mr. Skerrett: Does not Professor Mclntyre’s formula eliminate the piston stroke? Witness: To make the formula com-/ plete the piston speed must be con-, sidered. . , Mr. Skerrett: What time did you allow for the valves to be open when calculating the velocity of the gas ? ' Witness: About one-fifth. His Honor: How do you arrive at one-fifth? , Mr. Skerrett: In order to determine the velocity of gas, have you not to take the volume of compressed gas, and the time during .which the valves were open? Isn’t it common-sense?

COMMON-SENSE AND ENGIN-

EERING

Witness: It may be common-sense but it is not good engineering. His Honor: Then good engineering is not common-sense ?

Mr. Myers: Good law is not always common-sense.

: Witness: To determine the velocity of gas being discharged from the cylinder, the proportion of time the valve was open should be determined. This was qualified by the fact that the speed of the piston was not the same.

Mr. Skerrett: Supposing the discharge valves were open during the whole of the stroke and the speed of the stroke was uniform, what would be the velocity of the gas? Witness: 173 feet per second. Mr. Skerrett: I want you to do a small calculation, and I will show the schoolboy error they had fallen into. He asked witness to assume the same conditions as above, but assuming the velocity was 32 feet, what do you get, multiplied by 11 and divided by 2? Witness: 176.

Mr. Skerrett: In arriving at the period of time during which the piston had travelled, have you not taken 2-llths of the stroke ? ,

Witness: Certainly not,

Mr Skerrett asked whether the proportion of time the valves were opened was not considered. Witness said it did ifot attempt to do so. The formula used was a practical one used by engineers. - Mr. Skerrett: I put it to you that you engineers have made a mistake of a school boy] you have taken a proportion of length and called it time. Isn’t that so? Witness: 'Certainly not. Continuing, witness said the volume of gas drawn into a compressor was a steady quantity, and was irrespective of varying back pressure. Mr. Skerrett: I understand you to say that the Linde machine at Nelson Bros, is not a standard size—l 3 x 30? Witness: I believe it not to be. I am conversant with all Linde sizes, and I do not know of any that size. Mr. Skerrett: Are you prepared to contradict me when I say it is ? Witness: 13 x 30 is not a standard size. They have not sold machines of that size for many years. He could not say whether this was a standard size. He had been told that Linde valves had been working on a. high revolution machine at Woodville. It was quite possible that a valve would work satisfactorily on a small machine but perhaps not oh a bigger machine. He was aware that the National valve was constructed on the same principle as Linde.

RESPONSIBILITY FOR THE VALVES.

To Mr. Myers: The diameter of a cylinder of a 36-ton machine was larger than that of a 10-ton, and the clearance of a valve on a 15-ton machine would not be sufficient for a 36-ton machine. The velocity of the gas would be increased if the valves were not open all the time and the same quantity of gas was put through. If the "valves Were only open 5-18ths of the time, the velocity would be more than 173, and it would be bad for the valve. In his calculation he had assumed everything in favor of Niven’s valve. The valves would he open longer in a machine of longer stroke. A larger clearance was not necessary with the larger piston? When answering Mr. Skerrett lie overlooked the fact that the valves would he open a longer time. It was rjpite possible that the accident would have taken less than hours. His Honor asked whether the shape of tlio lie a-cl conic! be filtered by light taps in less than hours. Witness replied that he did not think it would take only minutes to complete the accident -from beginning to end. , ' His Honor asked whether the consulting engineer would examine the valves before passing the machine Witness: No;, the responsibility would be with the contractoi. His Honor said that he could understand that were the valves standard but he could not with so many different varieties of valves. . To Mr. Myers: The engineer might specify the number, but lie does not specify the design in the valves. Judging by the suction pressure in the log-book, he did not think the machine was working at, full capacity. There was no means of knowing whether there was any fractured stops by the entries in' the log. The siiction pressure would he much heavier during killing periods. The marks on the piston could have been made during the first adjustment-.of piston clearances, hut it was quite probable that it was done during the breaking of the spindle. There was no evidence on the guides of there being a buffer to the valve and he did not • think it possible for the guides to have held the stem. Unffiss lie saw the design of Niven s valve, he could not say that he wquld know the extent of the machinist’s error. Mr. Skerrett said that witness had Bold him that he knew the extent of 1 Honor: ]■! understood him to say that the machinist s error stood between the valve being a good one. Mr. Myers: Provided the design o-ave sufficient clearance. . • Witness: I assumed the design,gave sufficient clearance, -, . His Honor asked how lie could compare the valve with the others it he hacl not seen the design. Was it due to the machinist’s error? , Witness: It might have been through the design or the construcMr. Skerrett 'asked whether the calculation was made by witness independently of others. Witness replied that it was.

EXPERIENCES AT TAIHAPE

David Rutherford, mechanical engineer. said h® was engineer in charge of the Otaihape works,-near Taihape. ,Ho had' had experience i n some of the big companies m England and America/, and has had ten. years’ New Zealand /fexperience., ll® had been employed at the Wellington Meat Works beforo joining the laihape works. Last week he had 1 examined tlie broken-down compressor. He accounted .for., the valve getting into the cylinder to a general fractore, and - from his , experience tlhe fracture was due to the design of the

valve. The design ,of the valve was not adapted to the velocity it had to run. He wa s quite satisfied that the clearance was not sufficient. At Taihape there was a 24-ton compreSh sor. There were three delivery and three suction valves at each end. There was, a break-down through a mushroom head falling into the. cylinder. He had not known a similar experience in the other types of .com'pressors lie had operated. The accident happened on two .occasions. On the first occasion, Niven’s men were in charge. It was about a fortnight after it had been started. Witness had watched the machine previously. The accident happened about midnight. Niven’s engineer was running the test, and one of their representatives was there. The valve got into the cylinder and • the machine was stopped to see what the ■break was. The mushroom head was picked out when the machine was opened up. The mushroom head was flattened out to about three parts its thickness, and the general appearances were much the same as the mushroom .head, (produced) and which come from the Poverty Bay Works. The stop and valve spindle was one piece, and the valve was not designed! in the same manner as the Poverty Bay valve. Tho spring was broken. After .witness took over the plant, about three weeks afterwards, .a similar accident occurred. The valveliead was similiaa-ly pounded, the spring broken, and the remainded of the accident presented similar appearances. He had been given no previous warning of any /kind. When this compressor was working the beat of the valves was different to those in the slow revolution machines, and lie described it as a tap with a break in it. 'Previous to the bi'eak-Cown, witness heard no sounds different from what he liad already heard in tlie maenine. Later on he experienced a third broken valve. He coulcl put no faith in the tune of the machj ine. He had had all the valves out and overhauled a day or two previous to the third valve breaking. A mushroom head and spring had broken but the head did not get into the cylinder. He had a suspicion that' something was wrong and examined, the valves, finding one broken. He, after this, ordered the valves to be examined twice a week to make that there was no break. He was also making experiments trying to reduce the travel of tlie valve. In addition, lie bad occasionally examined the valves and occasionally came across a broken spring. In other compressors lie had never come across a broken spring. In England it was the practice to overhaul once a year in the coldest part of the year. This was also the case in New Zealand.

INSPECTION OF NEW MACHINES

To His Honor: In a new machine ho would examine the valves once or twice a fortnight. Hie third break occurred a couple of months after tlie second. Tlie area of the clearance of the valves of-the machine per valve was 1.8653 square inches. The total clearance of the three delivery valves -was 5.5959 square inches. There was a gradual change in the valves sent by Nivens to witness’ ol’derMr Skerratt said the 24-ton compressor was fitted with Linde valves. His Honor: Then we have been barking up tlie wrong tree. Mr Myers: Oh no, Yoar Honor. Mr Myers Avas producing a valve cage— His Honor: An’t we getting'mixed up? Is that the broken cage, Mr Myers ? Mr Myers: Yes. Your Honor. His Honor: Would you mind pulting a piece of red tape through it so it Avon’fc get mixed. Witness said the vah r es which broke doAvn Avere practically the same as the Linde valve. It bad a cage ..similar to The cage in the Poverty Bay Works.

Mr Skerrett said that Linde valves Avere fitted to 24-ton You are barking up the wrong tree. Mr Myers: Oh no, Ave are not. Witness, proceeding, said that the first alteration which took place Avas the separating of the permanent stop and thickening the spindle, in an endeavour to eliminate the break. This did not alter matters. The third valve Avas a lighter one machined out at the fillet to reduce the v'eight of the vah’e, tlie object being to reduc© the excessive Avear on the machine. In seven months’ working eight valves had been gone through, three breaking. The compressor Avas not run continuously Avithout a stand-by. After the machine bad been fitted with the neAV type of valve it had given every satisiaction. His Honor: It is tlie first valves avo have heard doing so. Witness said the neAv A'alves ha<t cages similar to the Linde cage, the' valA'es Avere similar except that the stop Avas separated. The ratio of clearance to piston area in the first valves fitted to the 24-ton compressor Weis 5 per cent. The urea of clearance for the present valve Avas I--565 square inches and the ratio of clearance area to piston area was- 6 per cent. He did not think from his experience, that the Poverty Bay engineer had received* any Avarnmg of the coming break. It Avas a bal’d thing to say.

•HOBSON’S CHOICE.”

/ To Mr Skerrett: The first comprossc>r put in was a 24-ton. The valves put ill were supposed to be Linde valves. He was not able to contradict the assertion that. the. valves fit-, ted on the compressor were Linde standard valves, manufactured by Linde and imported by Nivens. lie first mushroom head fell into wie cylinder during the progress of erection. It was a ■■(busy tune and both engineers and erectors were pushed on because the company wanted to 1 start killing. Witness had never previously worked a high revolution compressor. A Mr. Hein was m charge of the compressor when the muchroom. head fell in. The second valve which, fell in was exactly the same type as fell in the first time. On the third time the type of valve was the same, but the stop was cu so as to reduce tlie lift of the vali e. Witness reduced the lift of the Linde valves from 3-Bths to between a i a.nd 5-16ths. He commenced the experiments aliout April and, continued until September By reducing the Hit of the valve, the area of clearance was reduced. Mr Skerrett: Then it was not insufficient clearance which was the trouble, because /thelift had been reduced. There was apparently nothing wrong with the valves as far as clearance was concerned. Mr Skerrett: Did you stop up a discharge valve ? y Witness: Yes, How did it work.? ■lt was Hobson’s choice.'' It worked apparently all right! Did not the closing up of the valve show the clearance ample? No. Witness went on to say the delivery pipes were not unduly warm. ■Before the new cover was fitted, springs Avere continually breaking, but not afterwards. He didi not know that Mr; McCarthy had suggested putting in t-lie same size of valves when the compressor w»3 converted. Mr Skerrett: . The new valves were No. 5 Niven, were they not? Witness: I won’t admit that; there has been too much said about the Linde valves. • They were Niven’s valves, were they not? Witness: I won’t say. At this stage the Court was adjourned until 11 o’clock this morning. ~ ■' '- ' ' '