Vintage Tappets

[k. rogers]

Can somebody please tell me the best method to set up valve clearances on a '29 non-detachable cylinder head engine. Am I correct in thinking they must be set when hot? And does 3 thou sound right - that's the figure I have found in the technical pages unless someone knows better!

[Tony Fearn]

Hi Ken.
It might be worth giving Mike K......h (previous tech. info. officer) a ring. His vintage engine set ups go like the proverbial.
Tony.

[k. rogers]

Thanks Tony, but this evening I found the answer in an instruction manual - the obvious place to look! They should be set at 3 thou when the engine is hot. At the moment I can get a 3 thou gauge in when cold so you would think they will get noisier when hot, but they are still reasonably quiet! My original concern was that they had enough clearance so I will probably leave well alone at present.

[Keith Clements]

My principle when setting tappets is to have as little clearance as possible as that reduces noise and increases the amount of charge you put in and exhaust you push out. There is another requirement of cam design in allowing time for the heat to transfer out of the valve into the head, but I have not found this a problem on the low revving Jowett engine . Having less than the recommended tappet clearance or lash will reduce this cooling time. Having larger than specified tappet clearance or lash can severely damage the camshaft or valve train.

The limiting factors are to make sure the valve closes tightly under both hot and cold conditions, there is sufficient time for the valves to cool and that there is not undue strain on the valve gear. Since the block, push rods and valve gear are all made from different materials, and geometries change between different designs, predicting the theoretical effect of temperature can be an engineering problem. In high revving engines, dynamic considerations may also change the optimum clearance.

The general principle is that exhaust valves get hotter than inlet valves, so they will need more clearance as the valve stem expands under the intense heat so opening the valve when it is supposed to be closed. This allows exhaust gas to flow past the valve further heating it until burnout. Setting when hot (not just warm) is not practical as they will cool down from max operating temperature quickly. However, when the book says hot it means when the coolant water or engine block is at normal operating temperature, not when the valves are at operating temperature. It is probably more accurate than saying 'room temperature' that is cold, but it could also be that for a performance engine you are not worried about what happens when cold. Some operating manuals say set when hot and some, like the Javelin and Jupiter, say set when cold. I have yet to work out why this is for a similar design such as horizontally opposed, aluminium block, iron head. Note also that cold means when the engine has been switched off for at least 6 hours and presumably with the block at about 17deg C or colder since the cars came from Yorkshire in the early part of the last century!

How much you can safely deviate from the recommended lash settings depends on a few things. If you are using a solid cam that recommends </= 0.015" hot lash, you should not add more than 0.002" (two thousandths looser), or subtract more than 0.004" (four thousandths tighter). This is because of the lobe ramp shape- if you add even a little too much, the lifter hits the lobe away from the clearance ramp and will 'dig in', causing excessive wear in short time. Cams having bigger lash recommendations can go up to double the lash settings above in most cases, but be aware that there are many variations in how any particular engine is designed and built. These variations can have a direct influence on how much difference there is between hot and cold lash settings, and for how much you can "get away with" when going outside the manufacturer's recommendations!
Too tight is by far the lesser evil than too loose, as far as if you're going to make a mistake. Too tight will be apparent by a loss of power caused by the valves not seating and can overheat the valve because of a too-short seat time that doesn't allow the valve to transfer heat into the head seat.
But too loose will beat hell out of the cam and lifter as well as the rest of the valve train, because the take up ramp has been removed from the lobe. The acceleration will be too great for the valve train to be able to control; as much as a 40% increase in velocity has been seen with a lash setting just 0.002" past the end of the cam's take up ramp!

In aluminium block engines valve clearances need to be greater (if set when hot) as aluminium expands more than the steel of the pushrod (unless you have 'dural' rods like my Jup engine). One source says the expansion difference is about 6 thou from hot to cold with an aluminium block and iron head (but see my calculation below which reckons on 12 thou). The same source says 12 thou for an aluminium head and block.

On heating, the relative length of the pushrod is reduced as the block has expanded more, effectively closing the valve. Conversely on cooling from a too tight warm clearance setting, the valve will open causing blow by. Thus if the desired amount of clearance is .010" when warm you should set to .004" when cold, On overhead cam engines there is no such effect.

valve, say nimonic 80a, 12.7 µm/m-°C
head, aluminium alloys, 23-26 µm/m-°C

If you want to do the sums work on an exhaust valve head temp of 800deg C. but down to say 100 deg at the other end (average midpoint rise from cold (zeroC) to max 450C) with a valve length of 7cm and use a push rod length of 30cm and temp change of 80C with a difference of coefficient of expansion of 12 micrometres per metre per degree C. I worked this out to be the valve lengthens by 15 thou and the push rod shortens relatively by 12 thou. The inlet valve sees much less temperature and lengthens by 7 thou. This makes a theoretical cold setting (zeroC) of 3 thou for the exhaust valve and zero for inlet, assuming the block gets to 80C before the valve gets to 450C. If you run with a cold engine but hot exhaust valve then a 15 thou clearance for exhaust and 7 thou for inlet might be appropriate. These are very approximate calculations just to get a proof of concept. The actual lengthening of the valve stems may be less as little of the stem is exposed to the exhaust gas. These temperatures may only be reached after prolonged high speed running. But also bear in mind that valves do lengthen over time. VW recommend checking every 3000 miles. Jav /Jup valve setting by the book is done cold and set to 2 thou inlet and 6 thou exhaust.

Unfortunately feeler gauges are not very good at measuring clearances because surfaces may not be parallel or flat. Also humans are not very consistent at deciding when the feeler gauge can just move. So my method is to just make sure the rod rotates when at the fully closed position on the cam when cold (which might be at minus 10C if you want the engine to work on start up in those conditions).

Finding out when the cam is fully closed is also a challenge. This is best done with a dial gauge as you cannot be sure where the cam sits in relation to the crank. but various 'quick' methods are proposed by different experts. The dial gauge on the rocker will tell you when the rod stops moving up (closed) and when it starts moving down (opening). Halfway in between these two you should check the clearance. Note that the valve closed and opening angular reading might differ for each valve depending on the profile of the cam and wear on the lobes or the cam drive mechanism or (as shown by Philp Dingle) bend in the whole assembly due to dynamic forces and bearing wear.

All this is very technical and not what Ken needs as most people are happy to set by the book which always errs on the conservative side with larger than necessary clearances. On the Alaska trip we checked valve clearance every 4 days (1200 miles) mainly because it was a recently built engine but also it reduces the risk of a burnt or dropped valve.

[Keith Clements]

Another view from the VW camp which has a similar engine design and they have experimented with aly and steel pushrods.

The only reason we need lash is so the valve always fully contacts the seat (this cools the valve off, heat transfers to the valve seat where the head moves it the fins). The valve temp doesn't care if lash is .000 with clearance, or .040".

Now, the valve tip is another story. Too much lash and it will mushroom from the pounding (and you lose duration from too much lash too).

The VW factory used Al pushrods because valve clearances were more consistent from cold to hot. Once you use steel, all that goes to hell, since steel has a much lower rate of expansion compared to aluminum. Hydraulic lifters adjust for these changes, and if you have Hydraulics, you can run any pushrod you want .Back to the beginning: VW used to recommend .004 in and .006" exhaust clearances. They changed this to .006/.006" because lazy VW owners weren't getting their valves adjusted at the recommended intervals, and their intakes were going tighter then .000 clearance, and they were burning valves. Since you can't change customer habits, you change the spec to .006", which gives them 50% more time before they are at 0 clearance.

Now, early engines used .010", but that's a different deal due to rocker studs, I won't go there.

The reason we need clearance (besides the heat transfer deal) is because the valve stretches over time (the spring is lengthening the valve). Also, the valve pulls into the seat, the seat pulls into the head, etc. All this is worse the more spring pressure you have. Stock springs are easy on valves, Chevy or VW triples are harder. Lastly, if you run 9K RPM, the valves don't like that as much as a 4500 redline. Get the point?

All that being said, there's nothing wrong with running .000 clearance, but you damn well better check it often. Most of the CLF gearheads check their valves a LOT because they have a lot invested in their engines, and to not do it is $$$. I regularly ran .001 on my turbo engine with chromoly, never a problem and was much quieter then .004 or .006" (which was closer to .030" when hot BTW).

So, Al pushrods are the best, but until recently, they were too soft for serious performance use. Recent metallurgy has provided some killer heat treatments and alloys, that solve the problem for everyone, but they are about $100 a set instead of the $45 for chromoly. Al is QUIET at all temps, and due to this you can run .006" for valve safety, and have a beer instead of always crawling under every 3-500 miles to adjust the .000-.001" chromoly PRs, I have better things to do with my time!

Hope this clears this topic up a little, these are my opinions, you are entitled to your own.

John
Aircooled.Net Inc.

[Tony Fearn]

They should be set at 3 thou when the engine is hot. At the moment I can get a 3 thou gauge in when cold so you would think they will get noisier when hot, but they are still reasonably quiet!

Hello again Ken.
Am I missing something? If the gap is 3 thou when cold, then if things heat up, will the gap not shrink as the various metal parts expand with the heat?
Tony.

[Keith Clements]

This really depends on the materials of the valve gear and the block and head, the temperatures experienced by the components and the geometry of the system. If everything was made of materials with the same coefficient of expansion then the equations are much easier as all you are considering is the expansion of the valve stems which will tend to close the clearance gap and eventually open the valve. However, the aluminium block expansion (which is more than the steel push rod ) will counteract this closure since the push rod gets relatively shorter.

The key point is that if you heat everything (made of the same material) up to the same temperature the gap should remain the same as everything expands at the same rate. If you heat one component to a higher temperature it will expand more (this is how you get a starter ring on a flywheel). If you heat two components made of different materials up to the same temperature they will expand differently e.g. bimetal strips in thermal cut outs. Tappet clearance is there for two reasons. 1. to keep the valve still closed after it has been heated to 800+ degrees by the exhaust gas and 2. to take up any recession due to wear or stretch over time. The latter can be ignored if you check clearances frequently.

Crudely, the inlet valve clearance is there for lazy maintainers and the difference between it and the exhaust valve clearance is there for the exhaust valve expansion at normal exhaust temps.

When you set the clearance when the engine is warm (at say 75c) you are ignoring the differential expansion of block and push rod and only worrying about valve expansion when it is very hot. By setting when warm you run the risk of having the valve open up when the engine first starts. Thus loosing your compression. You might choose to ignore this as most performance car tuners do as such engines warm up quickly. But if you are one who takes the thermostat out and runs the car in the winter cold this might not be a good idea.

The answer to the question has to be done by experiment to verify any calculations. It is fairly easy to check the effect of a rise from say 10 deg C to 80 deg C but far harder to check after a prolonged burst at max revs and load. A dyno or spark analyser (Crypton) would be able to see the loss of compression when the clearance went negative and the exhaust valve opened. You might also be clever and rig up some optical system to check it was happening (although engine oil might scupper that idea).

Like many things once the clearance goes negative things get rapidly worse as there is no cooling contact between the valve head and the head, there is exhaust gas rushing past when it should not and the mixture gets weaker , all of which conspire to higher temps and eventually the head coming off the valve and disaster.

But do not be lured to not fixing a noisy tappet where the clearance has increased. This will wear cam lobes and valve gear, break springs and bend pushrods as well as reducing performance.

In Ken's case, assuming there are different metals involved, he should be able to see a difference between the cold and warm clearance and I predict the gap will increase. As the engine is run hard the exhaust valve clearance will reduce. The factory has allowed about 2 thou for this, the rest being there for stretch and wear. The inlet valve may also reduce a bit as well. Personally I would reduce inlet clearance to 1 thou. In a pre-war engine I suspect exhaust temps are a lot lower. The vintage engine has less aluminium in the geometry so the differential expansion is small.

[Tony Fearn]

The key point is that if you heat everything (made of the same material) up to the same temperature the gap should remain the same as everything expands at the same rate.

Valve gear 1934.jpg

If the gap (just above the valve nut -item 15 in the scan) is 3 thou. when cold, and the steel valve tappet (AS 22A) lengthens or expands when heated, and the steel valve stem (2007x1) does the same, as will the cam lobe, then surely the gap will decrease??
Tony.

[Keith Clements]

Well no. Because the valve and camshaft are constrained by the head and block which, if made of the same material and heated to the same temperature, all expand together . That is the distance between the centre line of the camshaft and the face of the valve will increase by the same amount as the components in the valve train. The net effect is the gap remains the same.

The exhaust valve will expand more because it is hotter than the material by which it is constrained (the block) which is (one reason) why you have a clearance. You need to have a drawing of the whole assembly to understand the geometry and knowledge of the material coefficients of expansion of each component and the temperature changes in each component to predict exactly what will happen.
Think about it like this. If you have a flywheel starter ring and you heat it up, its inner diameter gets larger, not smaller. You might think that because its metal expands the diameter would get smaller, but it does not because the circumference of the ring is longer than its diameter and so expands by a greater distance. 2 x pi x r and all that

If you really wanted to be pedantic the valve train would actually expand less because you had an air gap there (the clearance) and so there is less metal to expand than in the constraining casing. So the gap would open up!

The gap would open further if there was aluminium in the block or head or copper in the head gasket (if there was one before the valve face).