Temperature effect on valve adjustments?

[Eagledriver]

These are interesting theoretical arguments. I have no doubt that the exhaust valve does get hot and expand. The valve lash spec is actually .1mm plus or minus .05mm (.002-.006 inches). I've seen this published in a book and on this forum.

-Andy

[psalt]

I'd content that the actual operating valve clearance in a running engine, as opposed to a "hot" stationary engine actually decreases with increasing temperature.


Are you suggesting that they start out small (cold), get tighter when really hot, then get larger than cold when only warm ?

All of the hot running clearances I've measured have been wider than cold. Have you ever actually measured any ?

[psalt]

The valve lash spec is actually .1mm plus or minus .05mm (.002-.006 inches). I've seen this published in a book and on this forum


Which book would this be, the factory manual and spec book I have just say 0.1mm ?

[jcge]

Paul - I'm simply suggesting the RUNNING lash clearance gets smaller and smaller with increasing temperature on a RUNNING engine, and that the component with the greatest influence on lash is the expansion of the valve between it's cold and operating states.

When you shut down a running engine, the valve loses its heat very very quickly, much more quickly than the heads/cams/followers, and I contend that by the time you get to measure the clearance, the valve would have contracted significantly, thereby INCREASING the HOT stationary measurement (as you and others have found) at the time the measurement is taken.

My point is that the clearance you are measuring, is NOT the running clearance, which is significantly less than the setup or hot measured (feeler gauge) clearance.

I've never measured the RUNNING clearance or HOT stationary clearance. I have no doubt the HOT stationary clearance exceeds the COLD stationary clearance.

Hope I've clarified

John

[psalt]

John,

No, my point about your theory is simple. How can the cold clearance get smaller when really hot and larger when warm ?

[jcge]

Paul

I'm saying that the increase in clearance that you measure between cold (0.1mm) and hot (0.25mm-0.35mm when stationary) is all but taken up by the valve head and stem expansion when it's running at 500°C+.

Are we on the same page ???
John

[Flieger]

But valves are usually stainless steel, right? The coefficient of thermal expansion for stainless steels is "nothing" compared to Aluminum. The combustion teperatures would be mostly touching the stainless steel so I think the difference between a running engine and a hot stopped engine would not be so great, as the Aluminum would still be "warm". The Aluminum would not see the swing in temperature. Valves, cams, and rockers should not expand much- most of the increase in lash would be the Aluminum cam carrier heating and expanding.

I am not sure about the thermal conductivity of stainless steel compared to Aluminum, or Titanium, for that matter. The stainless steel may hold on to the heat longer than the Aluminum.

It would be interesting to examine an engine with DFLs and ceramic/DLC coatings on the pistons, combustion chambers, and exhaust port liners (like the 964). That would block some of that heat and would be a good way to test your theory.

[psalt]

I'm saying that the increase in clearance that you measure between cold (0.1mm) and hot (0.25mm-0.35mm when stationary) is all but taken up by the valve head and stem expansion when it's running at 500°C+. Are we on the same page ???


John,

Your theory seems to be that there is a sudden, disproportinate, expansion of the length of the valve during running that closes up the clearance and disappears as soon as combustion stops ? And it is kind of like Schrodinger's Cat, if I want to measure it, it's not there.

My questions would be , how much how and quickly ? Since this stem expansion would have to be greater than the actual measured hot stationary clearance, why doesn't the same expansion take up the guide clearance and ruin the seat fit ? Around 75% of the heat goes through the seat not the stem and the stem tip would not be +500 C. I have stood next to engines on a dyno that have had the clearances checked immediately after shut down, so your expansion would have to disappear in seconds. I think this is a complex issue involving the net expansion of several components. It is somewhat counterintuitive that an air cooled engine would have such a small exhaust valve clearance spec. I have other engines with sodium cooled valves that are water cooled and they have much wider (0.020") exhaust clearances. Guide clearance spec however, is wider for the air cooled engine. I do not know of any auto engine with a tighter exhaust valve clearance spec that the Porsche factory spec of 0.1 mm, have never seen a range published by Porsche, and have never seen an OEM recommend diffferent clearance specs for minor changes in ambient temperatures. Have You ?

I think the main variable in the different clearances spec is to prevent closing up of the gap to zero in between the service interval.

Paul

[Flieger]

The steel valve cannot be expanding more than the Aluminum cam carrier or Aluminum head. The steel holds on to the heat longer, too. Valve lash immediately after shutdown should be the same as running at operating temperature.

When the engine is still, it is heat soaking since the oil and airflow has stopped. This causes temperatures to spike a little while after shutdown. Lash should be the same or greater as running, assuming lash grows with increasing temperature.

[camgrinder]

.004" cold to .012" hot is my belief. Here is a plot of a 906 intake cam at the valve. The green line is .012" lift and the blue line is at .004". The factory designed the "clearance ramp" long enough to keep the valve seating velocity low, hot or cold.

There is a 56 degree change from cold to hot going from .004" to .012" on the 906 intake. My guess is the engine really wakes up when it gets hot.

[psalt]

I found a reference for the expansion of a valve stem when running. It is from Philip Smith's "Valve Mechanisms for High Speed Engines", what I would consider a good source. I have all his book and they are not at all error prone. The expansion of the stem under operating conditions is 0.0005" (ferritic) or 0.00075" (austenitic). There would have to be an expansion in excess of 10-20 times this stated figure for the hot running clearance to be less than the cold clearance spec, using the actual, measured hot stationary clearance.

From the reverse engineering department, Crane cams recommends subtracting 0.012" from the MFG hot valve lash spec when assembling a cold aluminum engine, when the MFG does not specify a cold clearance. This is in the ball park.

And from the truly bizarre department, we have the Chrysler slant six. In this case, the MFG, and I must say in 1959 they were riding the crest of the wave that top fuel engine still enjoy, recommended setting the exhaust valve lash to 0.020" while the engine is running, a perfect test for this theory. Not having done this, I spoke to someone with +40 years with this motor and they confirmed that the running clearance varys less than 0.002" after the engine is shut down.

You can see this exercise in action at :YouTube - valve lash on my 73 dart

[camgrinder]

Originally Posted by psalt

The expansion of the stem under operating conditions is 0.0005" (ferritic) or 0.00075" (austenitic).

This is .0005" per inch right?

Its normal to see timing cards with an additional .002" on the exhaust side vs. the intake because the valves can grow.

On all aluminum sprint car engine, the .012" number is a good rule of thumb. I like to give these guys a hot lash setting, and tell them to start it up .012" tighter and check it when its hot.

[psalt]

Hello John,

No, in this case he is talking about the total expansion of the valve in relation to the guide clearance. The issue we are discussing is whether the heat of combustion causes the valve stem to expand enough so that the actual clearance when running is smaller than the cold clearance adjustment. My experience with measuring hot clearances suggests not. Most of the other engines I know have higher clearances for exhaust valves, in the 911 that is not the case, and may have more to do with the expansion rate of the aluminum, not the steel valve.

I can't really understand what your chart is saying, both lines look brown to me. Can you give me a hint ?

[Eagledriver]

Wow, that slat six information brought back memories. I ruined a set of feeler guages setting the clearance on a running slant six. They were my dads feeler guages and he wasn't amused! Setting the clearance actually wasn't that hard. The problem was I was stacking thinner guages to get the proper number and the thin guages stretched and deformed.

-Andy

[Flieger]

Originally Posted by psalt

And from the truly bizarre department, we have the Chrysler slant six. In this case, the MFG, and I must say in 1959 they were riding the crest of the wave that top fuel engine still enjoy, recommended setting the exhaust valve lash to 0.020" while the engine is running, a perfect test for this theory. Not having done this, I spoke to someone with +40 years with this motor and they confirmed that the running clearance varys less than 0.002" after the engine is shut down.

You can see this exercise in action at :YouTube - valve lash on my 73 dart

Must be nice not to have to worry about clogged cam spray bars, center-lubed cams, or oil line restrictors. Just take the valve cover off and start 'er up.

[Flieger]

Compare that with this: 4V Desmo Adjustment - Ducati.ms - The Ultimate Ducati Forum

Unfortunately, many of the photos are lost in cyber space, but here are a couple:




and a video: YouTube - cjayheff's Channel

Only the Italians could design a system where you must remove the cams and finger followers to change shims and valve lash.

Don't drop a valve (into the cylinder) if you are doing this with the heads on the engine.

[camgrinder]

Originally Posted by psalt

I can't really understand what your chart is saying, both lines look brown to me. Can you give me a hint ?

The up and down lines, the green one is at .012" lift and the blue one at .004". The graph lines are the brown ones. The arc looking one is the valve lift curve, the other one is velocity. Between the green and blue lines the velocity curve is constant at around .0005" per degree. I assume from this design, Porsche expected to see a .012" hot lash when the engine was warmed up and making power.

[psalt]

Only the Italians could design a system where you must remove the cams and finger followers to change shims and valve lash.


Oh, give them some slack, they make excellent castings, nice shoes and decent wine...... The British made engines where you have to remove the head and grind the stems to set the valve clearance (Aston Martin ) and we Americans made engines that the crank and pistons have to come out to remove the valves (Miller, Offenhauser). Nobody's perfect. If you like Desmos' look into Michelat's 1914 Delage Type S engine..... the French invented everything and then what happened ?

[Walt Fricke]

Bo: Thank you for starting (if inadvertently) one of the rewarding discussions we get from time to time.

John JCGE: The graph you posted of valve temperatures is interesting. And looks like it is drawn from measured and slightly scattered data, then with a curve fitted? Seems logical: coolest on head near edge where heat can leave via the seat, and on stem hottest where farthest from either guide or seat cooling. Any more info on how this was derived? Just curious, as beyond quantifying heat at the valve and its distribution (valve closed? open? average?) I'm not sure just how it fits the lash discussion.

John D: On your graph, what are the units on the vertical coordinate? Zero something is part way up?

One trace looks like it is the cam lobe (nice bell curve). What is the other?

On the cam trace, the angle (tangent, differential) at any point is the acceleration?

Am I correct that more lash = less valve open duration? So the 906 would wake up because there was too much duration (overlap?) until hot? Would spend less time at the early and late small openings with lower flows, and more in the fat part?

This may be helping me visualize part of Paul's point: as the lash opens with heat, the valve stays closed longer (thus cooling more). The low angle of this ramp area means a small change in lash leads to a relatively large change in seat time?

I still don't quite understand why you can't do with the grind shape what you can do with altering lash as far as seat time is concerned. What good is the part of the ramp the rocker/follower never sits on?

Paul - As always, you inform these discussions with varied experience most of us (me, anyway) don't have. I like the idea that valve lash specs are related to anticipated service intervals. Having ground some seats and valves, I know that lash reduction due to wear (more like pounding) in these areas is certainly real.

Your post gave me a flashback to the last time (now long ago) I adjusted valves on a V8 with the motor running. Was easy, though depended on sound and not measurement as I recall. Can't hardly do that on a 911. For that matter, how would one quantify running lash? Something stroboscopic? And indirect, equating seat/open time to lash?

Same Smith of Design and Tuning of Competition Engines?

Walt

[psalt]

What good is the part of the ramp the rocker/follower never sits on?

Same Smith of Design and Tuning of Competition Engines?


Hello Walt,

Less acceleration and noise, like the quieting ramps that Jaguar developed in their parabolic cams to reduce noise at double and triple the earlier non parabolic cam clearances. Not a performance issue, a marketing response to the hydraulic lifter.

Same Smith, his +200 page tome on the poppet valve has been out of print for a while. I had to buy a zerox copy from a peg legged pirate 25 years ago. It has every wacky idea tried until 1967, and gets into the valve temperature and expansion rates of the sodium filled valves of the last of the highest spec piston aircraft engines before the jet took over. There is some recent work trying to understand the heat transfer at the valve seat, the finish of the surfaces, the mysteries and the mechanisms involved, but you can find a similar discussion on page 800 of the "Modern Gasoline Automobile of 1916". This book is truly amazing and I'm sure Harry Miller had one in his pocket.