Cylinder Head torque.

[a101960]

I have just had my A10 aluminium cylinder head professionally  overhauled, and it should be back from the Cylinder Head Shop very soon. Can I venture to ask, how important is it to re-torque the head? And how many miles after the initial torque down should the head be re-torqued, if it really is absolutely necessary?  As you are probably aware, there is a good reason for asking this question! The avoidance of the rather unpleasant task of removing the rocker casting and the subsequent refitting of it. I have a new solid copper head gasket. I know that there has been some discussion  on this site about the merits of solid versus composite gaskets. I have been advised that solid is best for aluminium heads, and either type is O.K. on iron heads. Have any of you ever fitted an A10 head and never bothered doing a re-torque? Incidentally,  Les Patterson, the proprietor of the Cylinder Head Shop recommended the following torque values: 3/8? torque to 25ft lbs, and 5/16? torque to 23ft lbs.

[RichardL]

A101960 (I'm thinking that's not your real name),

I am going to stick my neck out here, so that others with more experience can slice away at it. I am theorizing that the solid gasket would be better for alloy heads because the temperatiure of the head would not differ greatly from the temperature of the barrels. With the asbestos between the copper plates of the composite gasket, heat is not efficiently transferred. The aluminum will dissipate heat much quicker than the cast iron and, therefore expand and contract rapidly (by comparison with the barrels) as engine work increases and decreses. The heat transfer through solid copper is close to that of alumium, causing head temperature to be stabilized by the slow heat dissipation of the barrels.

As for retorquing, and this is where I am really guessing, the composite's insulator will, at first, be resistant to being compressed when the head is tightened down. Then, the fibers and binders in the material will be less durable after a lot of heating and will not "push back" with the same force as when the head was first torqued down. Now, I am going to stand back and let others say when and how often to retorque.

I think there is a whole other arguement about the composite being more pliable and, therefore , sealing better. I know others on this forum have said they have better success with sealing when using a composite, presumably, because the copper cannot be annealed soft enough. I am going to do a little research on annealing and, maybe, a little experimentation to see just how soft the copper can get. But that will be latter, as my rebuild gets further along.

Richard

[dpaddock]

Look at Groilly's posting "Oil Leak at Head Joint" five postings up  and you'll see a good treatise on this subject, including mine on page 2.

[groily]

It's a heck of a good question. In principle, and forgetting modern vehicles for the moment as they seem to be exempt from all this and need no maintenance ever, a head should be check-tightened within say 250-500 miles of (re-)assembly, to make sure that all is well, regardless of the nature of the gasket. I think most user's manuals of the day would make this point, and make it regardless of whether the said head were alloy or iron.
Like our friend a101960, I hate the idea of having to take off the rocker boxes - again - to do it, but I think it's a necessary evil, for peace of mind if nothing else. Solid gaskets may well be better on alloy than iron - and Manosound may have hit the nail on the head with his thinking as to why. However, composite gaskets these days rarely contain that excellent substance asbestos, preferring to rely on a mix of old cornflake packets and what looks like cobblers' glue. They can - and I have painful, really painful, experience with other vehicles - require several re-torques before they settle. Between each of which they lose many ft/lbs as they settle in. Solid gaskets will usually be safely settled after one check, and decently leak-proof if they were annealed before assembly. This is important - a new gasket may not be annealed and is hard as proverbial rocks - as has been pointed out in these threads recently by dpaddock, who obviously knows about these things. Annealing is a must, and is relatively easy - heat to dull red and then leave to cool in the air, or plunge in clean cold H2O. The difference is truly amazing and dead obvious . . . these chemist chaps, they're on to something.
As to torque settings, see previous threads. Thinking of the threads rather than of what what's being tightened down, I reckon a 3/8ths BSF cylinder head bolt is fine at around 34ft/lbs into a cast iron thread. Dpaddock talks happily of 40, and finds that most satisfactory, for an alloy head what's more.  But hey, 30-40 ft lbs, either sort of gasket, and with careful assembly there should be no problem. We have to remember that in their day these things were regularly maintained with primitive tools by people who'd never even heard of a torque wrench and couldn't afford a socket set. . . they worked then, and they'll work now!

However, I'd say the reason I've literally just spent several happy hours playing with the top end of my A10 is because the head WASN'T check-tightened after the motor was thrown together by its previous owner/seller. Can't prove that, of course, but I seriously do believe that a check tighten is a must for any re-built engine, despite the pain-in-the-idiot factor. Groily

[groily]

Addendum. Having said in the middle of the night, when judgment round here can occasionally be sub-optimal, that heads need re-torquing, decided I'd better follow own advice this am to see what, if any, slackening had occurred in some 300 kilometres since refitting head last weekend. Iron head, annealed copper gasket. First thing that was obvious was that the rocker box had settled a tiny bit, we're talking fractions of a turn on nuts and bolts. I used the gaskets, which may be why, and maybe I should just use sealant as proposed somewhere hereabouts. But nothing either abnormal or to worry about - shall simply check tighten again, and check valve clearances, after another couple of hundred miles.

However, to the head . . . torqued last Sunday to 34 lbs . . .
Not one of the fasteners had so much as blinked. Not even my home-made stepped stud and deep-nut. All still spot on. Most satisfactory, as dpaddock says of his alloy head, especially as there are no leaks, no nothing nasty. I don't think it would have been any different if I had left it a few more miles. Next weekend it's going to have to do maybe 350 kilometres so I wanted to have it checked ahead of that jaunt.

So. Does that mean it was a waste of time per the original Q, and that I could have saved myself the hassle of playing with the pushrods again? Arguably yes. But I wouldn't have been happy. I think the fact that the gasket was well-annealed is maybe responsible for this happy state, but hey, I don't really know. I'm just pleased. I shall now leave it in peace until the next hissy fit. Groily

[RichardL]

Gents,

Perhaps this will be an interesting aside on the whole question of torque wrenches versus experienced hands. Doing a little research, it appears that the fisrt commercially available torque wrench in the U.K. was released in 1942. Japan's first was in 1951 and, in the U.S. I think it was around 1946 (I couldn't find an exact date). Considering this, it makes sense that A7's and A10's may have been too near the start of the technology to be onboard. This may be why we can find torque ratings for A50's and A65's. Take a look at the paragraph below:

   In 1942 the "North Bar Tool Company" (as Norbar was then known), became the first company in Britain to commercially manufacture a torque wrench. The initital demand was driven by the need for the gasket-less cylinder head of the Rolls Royce Merlin engine to be accurately tightened. Bill Brodey and his partner Ernest Thornitt obtained a license from Britain's war-time Government to begin manufacture of torque wrenches and Norbar was born.

Richard

[groily]

Very interesting! You mean to say it was we Brits who came up with this! (Someone will no doubt tell me the guys were actually Scottish, as they like to claim the credit for everything invented there from the steam engine onwards - as my Scottish memsahib continually reminds me).
Amazing it's such a recent invention, but certainly explains why such new-fangled data weren't included in service sheets and manuals for toys to be bought be people very unlikely to have such a thing in the toolbox. Certainly my old Dad only got one in the late 60s, when he had his first ever aluminium cylinder head (on a Citroen as I recall). But he must have seen them when working on Spitfires (Seafires actually 'cos he was navy) in the war. Groily

[a10 gf]

Nice reading all this! Getting myself some torque wrenches took some time, and lots of damage, doubts and work to be redone before I got the point. I meet guys working on complicated engines without, from experience, but most relying on some share of luck. T-wrenches gives me security, peace of mind and lasting results.

The comment about heat transfer properties of a copper gasket makes big sense, never tought about that aspect, will definitely continue using copper. They can be reused many times, annealing is no problem and one can feel how soft it gets. A retighten is recommended in some literature I have read on the matter, and I do it after a while, even if it is some work (waiting for some day where wrenchwork-inspiration appears). Always found some slack before reaching a new correct torque. Of course surfaces must be in good shape and flat, but the copper will form itself and fill minor surface deficiencies.

I also believe the correct, little-at-the-time and even cross-tightening of head bolts is crucial.

e

[G/F DAVE]

Hi,Groily I just torque head down and leave it never had any problems. Just make sure gasket is well annealed & all surfaces are good (I always smear gasket with HYLOMAR  gasket cement around pushrod tunnel) just in case of oil leaks !! Most BSA A10 heads have been torqued up several times in their 50+ years some at the hands of bodgers so some are likely to be out of shape. If I suspect this I use the old method of glass sheet & grinding paste on faces as most engineering shops skim of to much for my liking.I have never used a composite head gasket on a A10 or any of my BSAs so cant comment but can guess its just the same torque down and leave alone. I do re-torque down my old M20s as they are prone to leaks if you dont, but the head bolts are easy to get to unlike A10s....... Dave P.S the BSA service sheets dont mention re-torqueing head once fitted....

[groily]

That's good to hear GF Dave and I might have been worrying unduly. It's just that I always do it on everything, same as I reach for a bottle opener about 6 pm. (And I did want to see how my home-made stud and nut were doing.) Haven't tried a composite on an A10 since I was a youf (when trouble was to be expected for lack of budget, tools and knowledge equally). But I'd try to get copper for choice too, which is what SRM sent me. Annealing, smear of grease (didn't think to add Hylomar, but good idea) plus gradual sequential tightening may explain why it's worked out nicely this time. But I promise I have had MEGA trouble with some composite gaskets, made without asbestos for elfnsafety reasons. (And funnily enough, my old Riley has the same general sort of rockers-in-box arrangement as a BSA and is also rather a pain in the A to retorque. Maybe I should get an M20/1? Or just continue to extol the merits of the parallel universe of AMC twins, as they don't have the fans they deserve!) Groily

[dpaddock]

I believe we need to review the basics re head gaskets and the purpose thereof, and see if we can justify our stated preferences logically.

Gaskets in general, and head gaskets in particular, are used to seal fluids under pressure between mating surfaces which, because of manufacturing tolerances, cannot reliably prevent their escape. The gasket deforms to fill the voids in these mating surfaces, thus avoiding additional manufacturing and handling costs. The higher the pressure, the more difficult it is to effect the seal.

The hierarchy of best sealing methods begins with manufacturing the cylinder and head in one piece as was done in the early days of internal combustion engine design. No joint, no leak.

Next came screwing the head to the cylinder. Welding and brazing were used, too.

Meanwhile, the use of gaskets continued to evolve but barely kept pace with rising pressures and temperatures. The lure of making an inexpensive leak-proof joint was compelling, however, and head gasket design kept evolving to fill this need. Note this gasket method required a comprehensive approach involving materials, surface preparation, fluid dynamics, clamping mechanisms, etc. And cost containment.

A gasket-less joint was the holy grail, a design which would permit easy separation whilst promoting good thermal expansion between the two surfaces plus other advantages, real and/or perceived. In fact, this method was near-exclusive in early days of sport motorcycling and was achieved by lapping the head to the cylinder. Its major drawback is cost - cost to prepare the surfaces, cost to handle and assemble the pieces, and cost to maintain the surfaces in subsequent handling.

Several contributors to this forum have touched on this matter, suggesting that the head and barrel be prepared on a lap using valve grinding paste and a sheet of plate glass. This is good, although why stop there? Why not eliminate the glass plate and lap the two pieces together, as mentioned above? Years ago I did this to a Jaguar XK 120 engine which was being prepared for a sprint car, using methanol and nitro at 16:1 compression ratio. No joint, no problem. Two motorcycle examples come to mind: the Manx Norton and the Gold Star.

But wait! you say. The Gold Star HAS a gasket. Well, yes, but its purpose is to seal engine oil draining from the head to the the push rod tunnel; combustion pressures are sealed by direct contact between the head and a spigot on the steel liner in the barrel. The Manx Norton, free from any environmentally-correct need to collect oil from the valves, simply provides spigot contact only.

OK. How about our beloved A7'S and A10'S? Well, they're provided with a perfectly adequate gasket design,  regardless of and in spite of the few naysayers who always seem to know better. (Really, did BSA become the world's premier motorcycle manufacturer by disregarding this important factor? Do they really want to challenge Roland Pike on this matter? Moreover, who of us actually had a factory head gasket "blow" in service?)

The factory gasket is an elegant design, even by today's standards. It has compressible rings around the cylinder bores and oil drains which provide simple, effective sealing for the fluids and pressures involved.
Compare these rings with the proven ability of the lowly spark plug gasket to withstand the combustion pressures encountered.

And the solid copper gasket? Unless it's fully annealed to dead soft AND the faying surfaces of head and barrel are super flat, the situation is not improved. The solid gasket presents almost twelve square inches of surface which must deform in order to seal against the pressures involved. This deformation will require a minimum of at least 35 to 40 ft lb of bolt torque. The standard laminated gasket requires approximately  half that clamping force. What about the supposed value of enhanced heat transfer? Well, why would you think that the engine needs such enhancement - BSA forgot about that when the engine was designed?

I've listed some of the facts. Now, let's revisit our own preferences and prejudices, unsubstantiated stories and inexact experiences and decide whether we truly need to modify our wonderful bikes.

Best,
David

[groily]

And even the very humble 2 CV Citroen (or the 2 I've ever played with anyway) has lapped on heads - nice. One of the best vintage cars, unless you're a Bugatti freak that is, to wit the 3 litre Bentley, had an all-in-one block and head (along with 4 valves per cylinder and an ohc, all in the v early 1920s). Much to be said for no gasket then, from both ends of the cost/quality spectrum. Just slightly awkward grinding in valves at the far end of a very long tunnel in the case of the Bentley.
But equally, much in favour of the gasket too. I've never tried to run an XK engine at 16:1 with no gasket, but have had a lifetime's amateur playtime with '60s XK engines, including 2 in cars that I still have. One is my daily mid-60s 4-wheeled driver (8:1 cr) and has had no mechanical hassle in over 10 years of moderate use although it looks a dog in all other respects; t'other a rather modified higher cr triple Weber carb job that produces about 50% more power from the same basic engine: never a hint of a prob with the standard laminated Jag head gaskets on either. Contrary to often ill-informed opinion, they are very durable motors - for their day! Just ignore the 16-18mpg fuel consumption though, which in and of itself is a limitation on mileage.
No idea about laminates on the A10 because I fitted what I was sent - ie solid - after thorough annealing. I'm sure both sorts work fine . . . but I do say that the want of asbestos in modern laminates where it was standard in the maker's originals, is the probable cause of regular hassle in maintaining the torque settings in the early miles, and I'm drawing on comments other than my own in saying it in relation to the tatty pre-war Riley I mention from time to time.
I have tended to extrapolate from that experience a preference for non-laminates, but it may well be ill-founded. Next time I'll use a laminate on the BSA and see if there is any practical difference. Hopefully it won't be any time soon! And I've never thought for more than a second about relative merits in terms of heat transfer, although there are probably some interesting points there . . . Groily.

[dpaddock]

Unless someone can show technically how and why the laminate gasket is inferior in this service, I'll just defer the matter but with the following observations:

1. How do we know they are laminated with asbestos? There are several other non-asbestos thermal fabrics available such as Cerafelt, silicate wools, Fibrefax, etc. which, like asbestos,  are thermal insulating materials.

2. Moreover, why do some extol  the virtue of enhanced thermal contact between head and barrel in light of the apparent fact that the OEM gasket was trying to avoid this? In fact, P.E. Irving cautioned about this in his epic "Tuning for Speed".

[RichardL]

d,

Not to speak for Groily, but I think he was noting that the modern laminates do not Use asbestos. I , for one, am convinced that asbestos was used in gaskets before its general banishment. I'm not expert in the chemical or thermodynamic proerties of asbestos or any of the products you've mentioned. However, does it not make sense that asbestos, by its reputation, may have been the most stable such product around in its day? I think the products you've mentioned are newer inventions. One might wonder if old asbestos was more thermally stable than the newer products. My original premise, when questioning laminates, was whether, or not, the fibers broke down over time and caused loose head bolts.

So, was P.E. Irving referring to heads having the same or different coefficients of expansion with repsect to the barrels/block? My premise on keeping alloy heads in close thermal contact with the barrels was that the slow cooling of the cast iron would stabilize the temperature of fast cooling the heads. It seems to me that performance would improve with uniform temperature in the chamber. Also, it seems that wear-and-tear on gaskets and head bolts would be less if the head isn't going through frequent and rapid expansions and contractions during changes in riding load. Roland Pike at BSA (someone I recently discovered) found that head bolts in alloy heads were breaking near the juncture of threads and shaft. He then had BSA change to head bolts with reduced shaft diameter to take the stretching due to head expansion. Is it possible they thought that the head should be thermally close to the barrels and, hence, could not depend on the laminate gasket to absorb the dimenstional differences? I haven't a clue. It is more likely that they found it much cheaper to make solid gaskets and didn't worry about expansion and contraction and heat differences until head bolts started breaking.

Sorry if this is long-winded just to hear an amatuer's point of view, but I do enjoy thinking about some these obscure engineering points.

Richard   

[groily]

Gawd, I don't know! All I was saying earlier was that sandwich gaskets with soft goo where once there was asbestos can require serious checks in the early miles. Technically, I have no idea what the gunge is that some makers use, but it doesn't appear as stable as the old stuff. I'll spare you a pic of a failed sandwich gasket made of some modern goo, but suffice to say the middle stuff had extruded all over the place - visibly round the water holes and oil holes and studs. Its replacement took 4 check tightens before it stayed put at the 34/5 ft lbs specified: after just 20 miles the nuts were at c 25ft lbs; after another 100 they were there again, after another 100 miles they'd eased to 30, and thereafter they have stabilised and I haven't checked in the last few months. Other owners/users report similar. Failure to go through the ritual practically guarantees a breakdown and the spark plugs . . . will be rusty.
That is absolutely not to say the laminates for BSAs do the same thing - probably made of different stuff entirely, whether the original part or pattern, from the long list of possibilities. As I said, I'll try one next time and see. What I can say is that the copper one I just put on hadn't eased off at all at all in the first couple of hundred miles (waste of time checking), has done another 500 since and was running like a bird when I took it out for an hour this afternoon. Oil tight, air tight, 'most satisfactory'.
And I am still not going to incriminate myself by commenting on thermal expansion and heat transfer! Groily