Massive Coolant Flooding - Cylinders And Sump

[PJGD]

I took a look at what might be involved in retrofitting metal o-rings such as Wills [UK] or Technetics Oflex [US] to give a robust head-to-liner seal:
Note that Wills recommend the gas-filled ring for automotive use because the gas expands when hot which improves the sealing ability. Also, because of the cyclical pressure loading, the groove needs to provide side-to-side support to the ring so that it does not fidget in the groove and thus wear out. It is important that the ring is crushed to fill the groove so that when assembled, the liner top flange is in direct contact with the head as shown.

Wills_rings.png

Essentially, my liner clamping scheme won't work for these rings because there is only room for three (possibly five) M4.0 mm cap screws and this will only give me a clamp load of around 2,500 lb (11,400 N), whereas the metal o-rings require at least 10,200 lb (45,200 N) clamp load to crush them.

If we were to rely on the existing head studs, then these theoretically could give a clamp load of around 31,500 lb (140,000 N) per cylinder which is much more than is needed and suggests that we might be able to get away with a lower torque setting in the 20 to 25 lb/ft range?

Although my scheme would not work with the metal o-rings, it might still work with annealed copper rings, so I created a drawing of the cylinder liner with a groove for such a ring. There is not a lot of room since the groove has to fall in-between the pitch circle diameter of the cap screws and still avoid the combustion chamber. I took the dimensions of the combustion chamber from the JCL cylinder head drawing and superimposed it on the liner; see below. Of course some heads have been modified over the years, so there is a risk here.

Liner-to-Head Registration.png

My liner is shown with five tapped holes for the (stainless steel) cap screws, but this is because the cylinders are "handed" and so rather than have two different liner part numbers, the plan would be to have one liner part number but with two redundant unused tapped holes.

Modified Cylinder Liner for Seal Ring.png

[Mike Allfrey]

Good Grief! I have triggered a topic that has gone to five pages!

Progress Report

The cylinder head for cylinders 2 and 4 was removed, the gasket did show small evidence of combustion blow-by at the cylinder swage-over rings, but coolant had not been leaking through. These two cylinder liners had extra 0.002-in. shims added when the initial head gasket failure occurred.

After cleaning up, the cylinder liner protrusion was checked. This was found to be at 0.005-in with the cylinder head torqued to 20 lbs. ft. as done previously. The nuts were loosened back a quarter of a turn each, and the gap increased to 0.007-in., it stayed the same after the cylinder head was removed. Both cylinder liners were within 0.001-in. after relaxing.

The decision was then taken to install the Nos 2 and 4 cylinder head, using new Bill Lock & Associates gaskets. The sealant used was Permatex Aviation Form-a-Gasket 3. As even a thin layer as possible was ‘painted’ on both sides of the gasket. The cylinder head nuts were tightened in stages, 20 lbs. ft., 30 lbs. ft. (twice) and finally at 35 lbs. ft. (again twice). At the second setting the gasket appeared to give a little, hence tightening twice at the second and third settings.

Tomorrow, the same procedure will be carried out on cylinder Nos. 1 and 3. It should be noted that this engine has the extended head studs through the coolant inlet ports. That is a modification which provides confidence should the cylinder liner protrusion be at top end of specification. For the new head gaskets used, it is felt that the protrusion can be 0.006 to 0.008-in.

It has also been concluded that, maybe, for the current Auckland gaskets, the protrusion could be revised to 0.008 to 0.010-in. or, possibly a top limit at 0.012-in.? Comment on that would be much appreciated.

After the other cylinder head has been installed, and prior to installing the valve rocker gear, a hot soapy water test with 9 psi air pressure applied will be carried out. the reason for not installing the valve rocker gear is to ensure that all valves are shut during the test. Of course, the sealant will be given time to cure properly first, twenty-four hours at minimum.

Even though I am quite exhausted, I believe I am getting somewhere. Grateful thanks to Phil Squire for his patience with his engine, that has to be admired. Also thanks to all who have contributed so far, that is greatly appreciated. More to come . . .

Regards,

Mike A.

[Mike Allfrey]

Progress Report

The second cylinder head was installed today and the sealant is curing. Soapy water and compressed air test due to take place on Friday.

Both heads have tightened solidly and, so far, no cracks have appeared.

Will report further later . . .

Mike A.

[Keith Clements]

Whilst waiting for my delivery of Hylomar I thought I would check the engine that was brought around for me to start. After fixing sticking valves and tidying up the heads, I checked liner height.
Oh dear, Not good. they should all be about 6 thou
3, 17, on 1,3
8,0 on 2,4

20201126_165312.jpg

We will never know if this was poor assembly or ageing.
I am fairly convinced it was the former as it was certainly a home job but until it is reset I will not know if there is any stretch in the block.
I could try putting shims on the top of the liner first to see if the block is flat as I suspect a couple of corners are lifting.
This also raises a thought that is putting the shims permanently here a valid option?
There is a risk the shims will move which might be mitigated with some very small dowels.
There is a risk of the shims allowing gas to seap between them which might be mitigated by sealing with a high temp glue.
I am sure someone would have bodged this way somewhere in the Jowett world!
I have previously cut a shim and fed it around the base allowing me to insert without removing the liner completely. I had no problems with the engine afterwards. Not recommended, but this was when I needed to fix rather than completely rebuild.

[Mike Allfrey]

Thanks Keith,

I am still in quite serious doubt with the engine currently being worked on. The main point about my concerns relates to there having been 0.006-in cylinder liner protrusion accurately set when the engine was assembled (second time around). I purposely used the lower end of the then recommendation, due to all of the welding at the cylinder block, and there having been copper spacer washers and shims installed. At this setting, the Nos. 2 and 4 crankcase did crack open between the No. 7 head stud and the coolant inlet port. This happened two days after tightening the cylinder head to 37 lbs. ft. Hence my wariness for opting for the higher limit of 0.008-in.

Like you say, I have been tempted to try shims 'glued' to the liner lips, mostly as an experiment - but I am too old now!

My prime concern with this engine is that the liners for Nos. 1 and 3 cylinders are still at the 0.006-in. protrusion previously set, which tells me that there has been no yield at the crankcase. During initial assembly, at the liners' ledge I applied the thinnest smear of Loctite 515 I could manage between the liners and their spacer washers, with no sealant at any of the shims. My understanding is that is good fitting practice.

The Bill Lock & Associates cylinder head gaskets have been installed and we intend to carry out the soapy water test as soon as possible.

Fingers crossed . . .

Mike A.

[PJGD]

Off on another tangent here but still with a focus on the cylinder liner and head arrangement, and yet still with an original Jowett connection once-removed!

Recall that David Hodkin [ex Cambridge University, ex De Havilland design office] was engineering assistant to Professor Eberan von Eberhorst during 1949 while the Jupiter was being developed at ERA (English Racing Automobiles) in Dunstable. After Eberhorst left to work full time at Aston Martin, David took over as Chief Engineer at ERA and was largely responsible for the magnesium framed G-Type ERA racing car which has a definite Jupiter look about its chassis (below), and was driven on occasion by Stirling Moss. For a few years after the Jowett contract, David used one of the prototype Jupiter chassis with a self-built body on it as a personal car.

The G-type ERA Chassis.jpg

Interestingly, during 1953, he worked on the design of a new racing engine with several relatively novel features. These can be seen in the attached graphic of the concept and include what can be called a "mono-head", i.e. the cylinder head cast in unit with the liner, either each cylinder individually [preferred] or all cylinders together in a more normal fashion. This mono-head module then slips into the crankcase with suitable seals where necessary, and then the DOHC sits on top of the head. Long through-bolts then securely clamp everything including the main bearing caps together. In this way, the bolts carry all the loadings, and the mono-head without a conventional head gasket can tolerate high cylinder pressures from supercharging without fear of seal failure. One wishes (and it would have been relatively easy to do so) that he had tried out this concept on his Jupiter from where some of the inspiration may have come.

ERA-Hodkin Engine Construction.png

[Forumadmin]

Interesting. Such a design must exist somewhere in production if it was viable. It certainly looks compact and light. How would it have been assembled and how could a V or horizintally opposed been built?
On another subject, I received a response from Bill Lock who hopefully will join this conversation soon. His reasoning for the lower liner height is that the gasket filling is a wire mesh, so less compressible and should give a seal on the water jacket between 2 to 4 thou. if it is larger, there is liklihood of too much stress on block.

[Nick Webster]

The engine described above bears a passing similarity to the "Sabrina" engine developed by Triumph circa 1959 for racing at Le Mans and subsequent development into a production engine. For various reasons the idea was dropped and a six cylinder development of an existing engine became the path followed. Subsequently, historians have looked back at the "Sabrina" with interest. Those involved at the time have related that the so called sandwich construction of separate layers all bolted together while useful in some ways was not at all suited to the economies of production engineering. As we know there are lots of good ideas that cannot make it past the accountants.

The "fixed head" takes us right back to the earliest days of motoring. In the case of the ERA racing engine described I would imagine angled valves must have made valve seat fettling interesting work. The last engine I am aware of with a fixed head was the Leyland 500 diesel fitted to the Leyland National bus in the early 1970s. It had a fearsome reputation for unreliability and was difficult to maintain without stripping it down entirely. A top end failure - of which there were many - usually smashed everything.

Regarding other features that have been discussed in this thread, I am reminded that many diesel engine manufacturers used an O ring seal on the liner skirts with a metal to metal seating of the liner - the Dennis O6 of 1946-61 had two with a bleed hole between so you could see if they were passing. The Foden FD two stroke of 1948 onwards was another with O ring seals. They also had head gasket failures although this was caused by heads overheating and warping. It was solved by an idea that has previously been mentioned. Individual heads were bolted to each wet liner and then the whole assembly bolted into the alloy crankcase, again with rubber o rings on the skirt and a metal to metal seating. Use of O rings therefore was hardly an idea original to Jowett, although perhaps unusual in the car world. It is worth saying that in general diesels run at least16:1 compression ratios and also will incorporate a degree of expense in the engineering that would not be countenanced in private car pricing.

It would be interesting to think if other car manufacturers were experiencing the kind of problems Jowetts had, which were coming on several fronts. In trying to get the Standard Vanguard engine up from 75 to 90bhp (7.5cr to 8.5cr) for the TR2, Standard Triumph was obliged to increase the head stud diameter as well as take the stud length right down to level with the base of the liners. And this with a stiff cast iron crankcase. Head nut torque is 105ft lbs! A similar change with the head studs would unfortunately require design changes to the Jowett engine that would alter it out of recognition.

So much of this we know well and has been chewed over before. Had the Javelin / Jupiter engine been given more time for development and dare I say it, experience, then many of the issues might have been anticipated. As it was, quite possibly given their history of engine design, initially Jowetts did not even foresee that there would be a need to improve an engine that at the time of introduction was entirely adequate for expected use.

[PJGD]

Nick,
I am almost in full agreement with you on all points! I am not quite so negative on the through-bolt sandwich construction, after all most HO engines use long through-bolts to hold the crankcase together, Javelin included. Most opposed piston engines [Napier Deltic, R-R K60, Leyland L60, Coventry Climax H30] are through-bolt engines holding it all together. The Rootes TS3 is another example. As you say, this can clearly be justified in these commercial engine cases, but perhaps less so in passenger car situations although in the late 1990's I was involved in the development of a 1.2L light duty car diesel commissioned by Ford Dearborn but designed by FEV. This engine, the DIATA, was a through-bolt design (I still have the drawings somewhere), but it was decided that a larger engine was needed so DIATA never made it into production.

The problem with the Leyland 500 fixed head engine was that the cylinders, cylinder block, and head were cast in one unit. Most of the time this worked as intended, but if on a cold morning, a driver applied full load before the engine had warmed up, the cylinders subject to combustion heat, which was constrained at its base [bolted to the crankcase] expanded axially faster than the rest of the cylinder block and tried to push the combustion chamber out at the top, in the process shearing the inlet and exhaust ports in the head allowing coolant to enter the engine. This of course resulted in complete engine failure. However, the ERA engine (and the bolted liner construction that I have suggested for the Jupiter) allows the liner to float lengthwise relative to the crankcase and thus avoids the Leyland issue.

Leyland Fixed-Head 500.png

Interesting that you mention the Dennis O6 engine; when I started my apprenticeship at Dennis Brothers they had recently stopped building the O6 engine, but there were still some sitting around in the test shop. As you note, almost all wet liner engines use o-ring seals at the bottom of the liner. The earlier [pre-war] O4 engine of which I rescued a rebuilt example that never got sold [and I think is now down at the Launceston Steam Railway museum], also has this feature. In the O6 cutaway below, 75 is the intermediate weep hole and 76 are the lower liner o-rings.

Dennis O6 Engine.png

On the topic of development issues that other UK car manufacturers contemporary to Jowett were having, this website goes into some detail of the problems that Armstrong Siddeley were having post war in spite of having far more resources than JCL. On this basis, Jowett did quite well.
https://www.designchambers.com/wolfhound/index.htm#MS

[Mike Allfrey]

Good Afternoon,

Well, at last we managed to fill the cooling system with hot water. After applying air pressure at 4-p.s.i. to the coolant, initially, it looked good, however, we found streaming soapy water leaking from the rear face of the crankcase adjacent to No. 4 cylinder. This is the welded area described in my technical note on the No. 7/10 head stud modification. Then we discovered a stream of water cascading into the sump area, adjacent to the tooling dowel sleeve, outboard of the sump gasket flange - if that makes sense.

Nothing for it, but to install another crankcase. Once the engine is completely dismantled (third time!), there may be a chance of a photograph.

This brings up a couple of points, the Auckland gaskets are capable of handling slight weld distortion (gasket surface sinking), the gaskets supplied by Bill Lock apparently, cannot cope with this situation. So we have learnt something!

Referring to a previous comment about Standard Triumph modifying their cylinder head studs to cope with anticipated extra power, the increase in diameter and the corresponding torque increase relate directly to the Ferguson TE-F20 diesel engine, also manufactured by the Standard Motor Company, so, tractor parts at full throttle on the Mulsanne Straight!!

More to follow . . .

Have a good Christmas and here is to 2021,

Mike A.

[Forumadmin]

Mike, I guess that could be viewed as good news.
The observation about gasket material suppleness is welcome. But I would not conclude that this means the less supple ones are inferior but just that they will not cope with a greater tolerance of gap. They may be superior for higher compression, more highly stressed engines. Also they might benefit from more goo (or a harder setting sealant) to take up the gap.
As has been reported here, many crankcases suffer from leaks even without welding. The added stress of welding probably opens cracks and needless to say if it had to be welded who knows if all the cracks were found in the first place. Crankcase leak testing , I believe, is a must before starting any rebuild.
Even then there is no guarantee that it will find all of them. You then have the difficult decision about whether to weld, use a sealant or internal coating, or just saVE for a next generation of owners WHO MAY HAVE BETTER RESTORATION TOOLS.

I use a piece of pond liner when doing the test so saving head gaskets.
Do not work on the engine on Christmas Day!

[Forumadmin]

https://jowett.net/forum/viewtopic.php?f=65&t=5367

[Mike Allfrey]

New Year's Greetings!

Today we started totally dismantling the engine. This was a startling exercise, because when the cylinder head studs were removed, it was found that the No. 4 studs in both halves of the crankcase were only finger tight. It was about the same with the upper rear studs (Nos. 7 and 10), which had been modified to pass through the coolant inlet ports. During assembly, all of the cylinder head studs were tightened home to 20 lbs. ft. and had been secured with Loctite Studloc 262 along with the appropriate primer. There were other studs that offered little resistance to being unscrewed, but yet others were held so securely that consideration was given to applying local heat to loosen the Loctite.

I have never seen, or been involved with a situation like this before. All of the studs were new manufacture using stainless steel stud material. In the past, removal of No. 4 studs has always been a real wrestling match!

This episode makes me feel incompetent, but I will have a chat with the Henkel people before going further.

Watch this space,

Mike A.

[Mike Allfrey]

Sorry Folks,

Either a typing error or memory lapse - it should have been Loctite Studloc 263, not 262.

Question - after weld repairs (extensive) is it possible that the crankcase yields at the cylinder head stud threads after the engine has warmed through several cycles? It is still very difficult to believe that the studs would loosen off after such a short period of use.

When the studs were drawn up as sketches for machining, great care was taken to ensure that each stud was seated in the crankcase on its shank within the counter-bores, which were of differing depths, due to previous abuse. One of the No. 4 studs needed to be 9/16-in. longer than standard due to the damaged thread boss.

With the replacement crankcase, extra attention will be concentrated during the head stud installation process.

More reports to come . . .

Mike A.

[Mike Allfrey]

Update:

All threads in the replacement crankcase have been checked and found to be in good condition. The crankcase set is currently being machined to take separate rear main bearing thrusts.

My thanks to all who have commented on this topic - it has been interesting to read.

Mike Allfrey.