Thermodynamics - Expansion of Motor Parts

[Ohio HD]

I started a new thread so not to take away too much from the Wood Lifter post. Below is that post.

Woods lifters fine scratches


Expansion of motor parts with heat. My understanding of thermodynamics is very limited, but I'm a decent researcher.

Larry has measured Evo motors in the past and has seen a length increase of 0.040". I have no reason to ever doubt what Larry tells me. I know he doesn't speak about things he doesn't have the answer for. I trust what he tells me.

One thing we have to realize is that the growth of the top end (we'll use either Evo or TC as they're designed very similar) due to heat, from the base of the cylinder, to the top of the cylinder head, and rocker covers for that fact, will expand at different rates. The reason is the entire top end is not the same temperature, and the density and type of material changes part by part.

I do think that the possibly of growth regarding the combustion chamber volume, and the clearance in the valve train will be limited to being a little less than the entire top end growth. I don't think that 0.040" is unreasonable, but I have always considered that 0.030" or less would be the number that effects combustion chamber growth. Just my own assumption as you would loose as much as a single point of mechanical compression with a growth more than 0.040".

Basic thermodynamics. take a 12" length of aluminum with unknown type and shape and dimension, other than length. The linear temperature coefficient of aluminum we'll say is 0.000023.
Then heat the aluminum to 325° from an initial temperature of 75°. The length increase of that part would be about 0.069".

Now take all above the same, except using steel, then the temperature coefficient changes to about 0.000012. Then the same temperature rise would indicate the length change to be about 0.036" for the steel part.

As I mentioned in the prior post, the top end can only grow as much as the steel studs allow it to grow. And the various parts of the top end will expand at different rates due to temperature and material type, and of course part density. Cylinder studs will grow at one rate, pushrods and lifters grow at another rate, etc., etc.

This subject of growth for the motor to me is very interesting, and unfortunately there isn't much documented information that's generally available. No doubt there are companies that have performed these types of tests in great detail. I wish I could come across some of this data as it applies directly to HD motors.

[JW113]

The cylinder studs will not limit how much the cylinders and heads expand. They will stretch accordingly. Consider the studs to be linear springs, holding the whole package together. As long as it doesn't expand past the elastic point of the steel, they will stretch. In fact, when the motor is cold, it does not have the final compression force on heads/cylinders, it goes up as the motor heats up. This is one of the reasons you don't want to fire up a Evo/TC cold, and give a bunch of WOT to redline blasts. You could blow a head gasket or base gasket.

-JW

[Don D]

As soon as you take one of the metals out of the context of the whole assembly the result becomes for that single component not a running heated (and cooled) engine. IE The aluminum cylinders with cast iron liners, pushrods and lifters growing with heat applied. Many metals, many expansion rates. Many years ago Blackhills Ken did a test on a hot twin cam and came up with something around .018" net growth, hot twin cam.

[Admiral Akbar]

Couple things. 

1. The round numbers you are using for coeffects of expansion are for degrees C and Not F.

2. It's not really thermal dynamics. It's really thermal statics or at least the numbers you are using re-static. You do mention as the motor heats which is dynamic but assume everything is the same temp.

[Admiral Akbar]

The cylinder studs will not limit how much the cylinders and heads expand.

Actually it will.  It's the stress vs strain stuff.. Youngs modulus.. It likely won't be much tho.

[Ohio HD]

Couple things. 

1. The round numbers you are using for coeffects of expansion are for degrees C and Not F.

2. It's not really thermal dynamics. It's really thermal statics or at least the numbers you are using re-static. You do mention as the motor heats which is dynamic but assume everything is the same temp.

Appreciate your jumping in.

I just took a look, you're right about the values being for C° and not F°.

I was using the simple example to only show difference in materials under temperature. I agree there is a lot of variables that come into play, various temperatures from cooler at the cylinder base to hotter as you go towards the head. As well the different materials, and their densities that effect expansion.

I have no way to even get to the root of how to calculate the various parts of the top end as to what they will expand, and how much or little the studs allow expansion, etc.

Don mentions that BH Ken once took some effort on a TC to see what expansion was. In my mind a value of 0.018" could be close to what possibly the combustion area may grow, or a little less. I need to dig more, see what I can find.   

[kd]

Couple things. 

1. The round numbers you are using for coeffects of expansion are for degrees C and Not F.

2. It's not really thermal dynamics. It's really thermal statics or at least the numbers you are using re-static. You do mention as the motor heats which is dynamic but assume everything is the same temp.

Appreciate your jumping in.

I just took a look, you're right about the values being for C° and not F°.

I was using the simple example to only show difference in materials under temperature. I agree there is a lot of variables that come into play, various temperatures from cooler at the cylinder base to hotter as you go towards the head. As well the different materials, and their densities that effect expansion.

I have no way to even get to the root of how to calculate the various parts of the top end as to what they will expand, and how much or little the studs allow expansion, etc.

Don mentions that BH Ken once took some effort on a TC to see what expansion was. In my mind a value of 0.018" could be close to what possibly the combustion area may grow, or a little less. I need to dig more, see what I can find.   

Admiral, glad to see you hopped in.  This subject started in another thread about lifters and was restarted here.  It comes up lots but is never really settled.  There's plenty of opinions. It could be a great conversation to put a few "suggested" expansion theories and dynamics into readers Digest terms and come up with a realistic number.

IYO - Is there any understandable way to calculate growth for example to determine the change in pushrod plunger stroke position (adjustment) as a typical sleeved aluminum cylinder grows?  This all taking into consideration the different metals in the barrels, (do the sleeves effect the aluminum growth rate, heads, pushrods (solid or adjustable) and ability of the cylinder studs to grow.  If not, what is a reasonable assumption that could be applied to something like lifter adjustment to abate hot start issues.  (I don't mean just throwing stuff out to see what sticks).

That's a big challenge but it is winter for some of us. 



Ohio, when I read Don's remarks I see it as confirming if the barrels and heads grow .0XX" the pushrods etc will grow also but not at the same rate if steel.  The difference Black Hills Ken found in his personal hands-on research between all  of the assemblies was in this case +.018".  I too think that sounds reasonable.

[Ohio HD]

:
Ohio, when I read Don's remarks I see it as confirming if the barrels and heads grow .0XX" the pushrods etc will grow also but not at the same rate if steel.  The difference Black Hills Ken found in his personal hands-on research between all  of the assemblies was in this case +.018".  I too think that sounds reasonable. 

To me that's the entire conundrum, determining what parts would expand how much. And under what conditions and temperatures, and how much are they retained by fasteners. 

[kd]

Exactly.  A set of solid pushrods will have less mass than adjustable for example.  The core of the adjustable will be influenced by the insert threaded portion and the stud.  Different metals may come into play minimally but also may need to be considered to confirm actual growth effect on pushrod adjustment.  I think exploring that relationship will be useful also.

[JW113]

The cylinder studs will not limit how much the cylinders and heads expand.

Actually it will.  It's the stress vs strain stuff.. Youngs modulus.. It likely won't be much tho.

OK, perhaps my understanding of physics is metallurgy is out of date. When (certain) metals are heated, the atoms they are made of vibrate more, forcing their relative distances to each other apart. So yes, when aluminum is heated, it will expand, in length, surface area, and volume. Note that I said 'cylinders and heads will expand', not grow in one dimension only. If you attempt to constrain that growth in one dimension, it will cause the metal to expand more in other dimensions. Like an inflating balloon trapped between to fixed surfaces. I agree, we are talking only thousandths of an inch, but still...

Back to cylinder studs. They are made of steel, which has a lower coefficient of expansion than aluminum. And thus why when the engine gets hot, the clamping force between the heads/cylinders/block increases. Yes, the studs will expand also when heated, but not nearly as much. They will indeed get longer, though, due to force on them from the expanding aluminum parts. This is by intended design, and why the (stock) studs neck down between the threaded ends. If the studs constrain the growth in length of the heads/cylinders, that then turns into distortion of the cylinder walls. That aluminum has to expand somewhere, right?

So correct me if I'm wrong, but I believe the stud's job is to clamp the heads/cylinders/block together without causing the cylinders to distort, not limit the length of them. Yes? No?

-JW

[14Frisco]

turns into distortion of the cylinder walls. That aluminum has to expand somewhere, right?

Is your "That aluminum" referring to "cylinder walls"?

Is the (inside) cylinder wall of the cylinder aluminum?

[Don D]

Not to discount the physics,
For clarification Kens testing was to release a pushrod and measure it just as lash became apparent, hot engine. So with a known preload cold VS preload determined hot the growth NET amount could be established.

How do you know or what is the growth of a cylinder that is die cast aluminum (242, 335, 332?? heat treated ?) with a splined cast iron inserted liner.

[Don D]

To answer my own question perhaps slightly I could heat one and measure in a lab oven that has reasonable temperature uniformity.

[JW113]

turns into distortion of the cylinder walls. That aluminum has to expand somewhere, right?

Is your "That aluminum" referring to "cylinder walls"?

Is the (inside) cylinder wall of the cylinder aluminum?

No, the inside is (I believe) ductile iron surrounded by aluminum. Even iron will distort if a force is applied to it. If this were not the case, why is it common practice to use torque plates on cylinders when boring them? I believe to "pre-stress and distort" them into a similar condition as they will be when the engine is fully assembled, yes? In fact, I've seen a time or three on Evo cylinders with bunch of miles on them, four lengthwise vertical "stripes" in the walls that correspond to where the cylinder studs are. Yes, even the iron gets distorted under the force of hot expanding aluminum.

-JW

[hrdtail78]

On the hotter side of things.  Head temp reading can be 320 to 350 F.  Put a cylinder in the oven at 350 and heat soak it.  Measure and compare difference from cold.  This can tell you how much it expands w/ out being held by through studs.  My tq plates have threads in the bottom plate.  Clamp head and cylinder and repeat test.  You can measure from bottom o-ring/ gasket surface to rocker box lower gasket on head.  Measure hot and room temp.

This could at least tell you if you are picking fly "Potty mouth" out of pepper.

[JW113]

How do you know or what is the growth of a cylinder that is die cast aluminum (242, 335, 332?? heat treated ?) with a splined cast iron inserted liner.

That is a great question, and given the complicated math of trying to calculate that, given the complex construction and properties of different materials, probably difficult to come up with a definitive answer. Empirical measurement is a valid way to get the answer. But on the other hand, why be concerned with it? With a back of the envelope calculation, knowing the CTE of the dominate material (aluminum), and what the maximum expected linear length expansion will be under worst case conditions, all you need to worry about is enough compensation range in the hydraulic tappet to handle it with some margin.

Or, enjoy the sound of tappet lash. Like from an Ironhead Sportster.


-JW

[kd]

Not to discount the physics,
For clarification Kens testing was to release a pushrod and measure it just as lash became apparent, hot engine. So with a known preload cold VS preload determined hot the growth NET amount could be established.

How do you know or what is the growth of a cylinder that is die cast aluminum (242, 335, 332?? heat treated ?) with a splined cast iron inserted liner.

If I understand correctly, that would mean the heated pushrod (first adjusted cold) was the measure of growth of the cylinder / head assembly.  I think it is generally agreed that the pushrod would have grown from the initial cold setting also.  Do you know if Ken determined the pushrod growth in so he could declare the "actual" cylinder ass'ly  linear growth from cold where the pushrod was first adjusted?
 

[Don D]

Kd, in honesty I don't 100% remember all the details just the concept.  It was a solid test, real world. Same could be simulated off an engine using torque plate bottom, head on top and hd studs and bolts. Then heat and cold test of a pushrod length.


"Yes, even the iron gets distorted under the force of hot expanding aluminum."

The stripes are from uneven expansion, the through holes interfere with the release of heat, hot spots. Harley fixed that on 110 & 117" and all of S&S cylinders. Just the tops have room for the head bolts. The center is reduced just enough for the studs to clear rather than a 9/16" nominal hole top to bottom.

[kd]

That's good information on the stud holes that I for one didn't know had changed. We have discussed the stud ghosted wear pattern before. I am wondering if it improved that phenomena but that will take the thread off track.

I think we are getting somewhere now.  It seems to me, the only real way to determine the different growth ratios is by (as you suggest) actually subjecting the components to heat and measuring the combined growth from "room temperature. That includes separating out the lifter / pushrod assemblies to determine their different dynamic growth from the cylinder / head assembly (with and without torque plate influence if possible).  For one, that should provide real numbers to allow a fairly close calculation as to the change in lifter adjustment.  We could even get into the weeds and include case deck height but that may be splitting hairs.  It may be no more than the change in adjustment that occurs in the thread pitch surface from locking adjustable pushrods with the jam nut.

[Don D]

I am tracking but the pushrods would need a similar hot and cold measurement

[hrdtail78]

Be hard to do accurately because all these parts are going to see different temps.  If head temps can get to 350f.  That doesn't mean the case see's 350.

Just did a data collection session on the dyno on an m8.  Head temp indicated 289.  Engine indicated 268.  IR heat gun by spark plug 237.  Case right behind rear lifters (trying to get where lifter bores are at) 187.

It is cold in my shop and I only have the dyno jet fans pointed at heads and no other fans blowing on the engine.

Since different engine components are going to see different temps.  Made of different materials.  Billet aluminum is going to expand more than cast for example.  For real world data.  Wouldn't the engine parts separately need to be heated to different temps?

[kd]

Be hard to do accurately because all these parts are going to see different temps.  If head temps can get to 350f.  That doesn't mean the case see's 350.

Just did a data collection session on the dyno on an m8.  Head temp indicated 289.  Engine indicated 268.  IR heat gun by spark plug 237.  Case right behind rear lifters (trying to get where lifter bores are at) 187.

It is cold in my shop and I only have the dyno jet fans pointed at heads and no other fans blowing on the engine.

Since different engine components are going to see different temps.  Made of different materials.  Billet aluminum is going to expand more than cast for example.  For real world data.  Wouldn't the engine parts separately need to be heated to different temps?

No doubt.  Your temp readings are a great example of why.  There are many influences such as ambient temp, highway or city riding, oil grade, oil cooler, etc.. I think the reasoning behind this discussion is more to agree upon the separate influences present and the resulting effects.  I don't think anyone is so disillusioned as to think there can be an accurate calculation.  I think the hope is to come up with a realistic dynamic growth measurement of the components especially as it may effect pushrod adjustment.  Complete accuracy will likely be impossible.

[Don D]

Your headed for a number that could vary 10% due to variables maybe?
Photos of the two style cylinders

[kd]

That's not too bad if you realize it going in.  Say randomly .040" with 10% variance could mean plus or minus 5% or .002 each side.  In the big picture that's pretty damn accurate.  That's a tough to make measurement to make every time on it's own so I would say insignificant.  This is building understanding as much as knowledge.

[Admiral Akbar]

Couple things. 

1. The round numbers you are using for coeffects of expansion are for degrees C and Not F.

2. It's not really thermal dynamics. It's really thermal statics or at least the numbers you are using re-static. You do mention as the motor heats which is dynamic but assume everything is the same temp.

Appreciate your jumping in.

I just took a look, you're right about the values being for C° and not F°.

I was using the simple example to only show difference in materials under temperature. I agree there is a lot of variables that come into play, various temperatures from cooler at the cylinder base to hotter as you go towards the head. As well the different materials, and their densities that effect expansion.

I have no way to even get to the root of how to calculate the various parts of the top end as to what they will expand, and how much or little the studs allow expansion, etc.

Don mentions that BH Ken once took some effort on a TC to see what expansion was. In my mind a value of 0.018" could be close to what possibly the combustion area may grow, or a little less. I need to dig more, see what I can find.   

The thing is that things grow at a different rate when using degrees F or C. If you are going the grind out the numbers, they might as well be right..  The growth will be time 5/9 for F.

I'll agree that the expansion rate will be different between top to bottom of the cylinder.  Maybe not that much tho. Cylinder head is what? Maybe 280 on the intake side?  330 on the exhaust? Engine oil temp is pribabably close to oil temp.. 220? I guess you could map it with a IR temp sensor.

I forget how Ken came up with the number but it seems reasonable.

What are you looking for?  Pushrod plunger change in the lifter?