Cams road test PW3 vs JS2

Would like to diddle some math to look at the various component growth (piston, rod, barrel) at temperature but need to know a few parameters to do it. Would anybody know the stock piston compression height (or other piston compression heights) and the barrel length from deck to deck? Thank you.

Would have to measure each engine's components d/t manufacture faults and prior machining. I was told by late Ken Augustine all Norton barrel tops are slanted .003" off from base for instance so depends on where ya measure until squaring up. Peels barrel indexed to base as well as head bottom to head steady flat. Would make a great science fair project to actually calculate each components thermal swelling in it particular expected temp zone and predict how close a squish ya can get away with per rpm. Then do some testing to see how much mechanical stressing and compressing going on and after than the mechanical bowing and bugling of parts like crank and cam and fastener stretch which i've actually seen and pray never again. Its all goes so suddenly non linear at some threshold so like creeping out on thin ice all is fine till ... Lacing a racing bicycle rim is like this, its best to get spokes as tight as the rim and hub can stand, which is learned by rim suddenly popping into umbrella shape so next rim done just short of that. There is over 7000 lb load on center of 750 piston crank nearing red line, which do ya think will deform more if that much and more was hung on crank from the rods, the rods stretching or the crank bowing between its end supports?

WZ507 said:

Would like to diddle some math to look at the various component growth (piston, rod, barrel) at temperature but need to know a few parameters to do it. Would anybody know the stock piston compression height (or other piston compression heights) and the barrel length from deck to deck? Thank you.

Click to expand...

Here's some data to start with:

Norton Piston Compression Heights

77, 78, and 81 mm Maney JE pistons – 1.625'
Stock 850 – 1.471” measured, 1.470 – 1.478 factory spec
Old style Maney Cosworth 81 mm, might have been machined by Steve for lower CR – 1.571”
Stock 750 – 1.540”
Norvil 81 mm – 1.510”

The only cylinder heights I have handy are from a couple I recently measured. I don't know what the factory specification is.

Stock iron 850 - 4.552"
Maney aluminum 920 cylinder - 4.552

Ken

If the cylinders are not square to the base, or the crankcase top not parallel to the crank axis .... ?

It occurs that the thermal expansion between crank mainshaft location and crankcase deck is also part of the picture. As the al. crankcase heats up it will expand to lift the cylinder and move the squish flat of the head a little further away from the piston top. The steel flywheel expansion amount from mainshaft to crankpin also is a factor. This amount would be subtracted from the case expansion to leave a net growth amount for that part of the assembly. The net growth amount will add to squish clearance.

So as far as thermal expansion goes, one has to stack up everything that is connected to the mainshaft and calculate that amount, do same for the crankcase top half and barrel, then compare the two numbers to see whether the squish clearance increases with temperature or decreases. This will all depend on the materials chosen. If the outside assembly is all/aluminium and the inside is steel , other than the piston, then the squish clearance ought to grow perhaps ten thou or so using rough number for distances and temp increase.
If on the other hand, the conrod is al. and the barrel is cast iron as with a stock engine, then it appears that the squish clearance might shrink a little with temp increase.

You will also need to know delta T from cold, which would be greater toward the top of both assemblies, where combustion takes place.
As mentioned earlier, there are other factors as work, this calculation is just the thermal/expansion consideration.



Glen

It pains me as a scientist and engineer to say this, but theoretical calculations are not too useful for setting minimum squish clearances in Commando engines. Minimum values, i.e. the point at which the piston is about to hit the head, are determined experimentally, usually at some cost in replacement parts.

Ken

Ken - thanks for the info.

As others have noted this is a complex problem that merits considerable energy to fully understand. That said, in my simple minded way I wanted to put an order of magnitude to the growth of the various components to see if squish clearance increased or decreased with heating, and to get a better handle on the magnitude of such changes. Addressing stretching and flexing of components is way beyond anything I could ever hope to figure out.

For my austere approach I imposed several assumptions to simplify the system which included the following. Dimensions used were for a 750 CDO engine. For thermal conditions, I simply assumed a starting temp of 70F and that all components reached an operating temperature of ~ 260F (smoking hot engine oil). Is this correct and do all components reach this temp? NO. Will this assumption grossly mislead us? I don't think so.

Next I segregated the components into items that would decrease squish when they grow, and ones that would increase squish when they grow. The squish "decreasers" that move the piston closer to the head when they thermally expand are - crankshaft, connecting rod, and piston. The squish "increasers" that move the cylinder head away from the piston/rod/crank assembly would be the engine case and the barrel. The cylinder head was not included here and a neutral deck was assumed.

I have no idea what the dimensions of the engine case are, but assumed that if I totaled 1/2 the stroke, rod length, and compression height, and subtracted from this the barrel length, the remainder would be the case height from crank center to deck, which I figured to be ~ 4.615".

I used the following coefficients of thermal expansion (CTE), where the average values are in micro in/in-deg F. These values vary a bit by alloy as well as by the temperature regime employed, but are close enough for this coarse treatment.

Material
Aluminum - 13
Steel - 7
Grey cast iron - 6

Results
The squish "decreasers" expanded at temp as follows, for a total expansion of 0.0201".
Rod - 0.0144"
Piston - 0.0038"
Crank - 0.0020"

The squish "increasers" expanded at temp as follows, for a total expansion of 0.0164".
Barrel - 0.0051"
Engine case - 0.0113"

Conclusion
The assumptions and thermal conditions described above would reduce squish by ~ 0.004" at operating temperature. This result suggest that although thermal considerations are significant as relates to squish, other mechanical factors (stretching/flexing, etc) are likely to be more significant considerations.

The above is quick and dirty play that answered a simple question that was on my mind. I hope my simple minded approach has merit. If I'm out in left field and the result seems implausible, I'd like to know that too.

PS - It's late here and things that might be clear/easy in the morning are very challenging at this hour.

Hi Ken, sorry to hi-jack , but when you said cylinder height is 1.552" , what do you mean exactly, just because I am looking for the common lenght , between barrel base flange and cylinder head top .........thanks for any help!

lcrken said:

It pains me as a scientist and engineer to say this, but theoretical calculations are not too useful for setting minimum squish clearances in Commando engines. Minimum values, i.e. the point at which the piston is about to hit the head, are determined experimentally, usually at some cost in replacement parts.

Ken

Click to expand...

Indeed Ken, experimentation is the only way to be sure with so many (unclear) variables.

In my own experience, Norton crank, alloy rods, iron barrels, racing use, revved to 7500 and occasionally higher; pistons would touch head if gap less than .040, if gap more than .060, evidence of combustion taking place in the squish band area indicated that the squish was starting not to work. So keeping it between .045 and .055 was my own conclusion (with that engine with that usage).

WZ507 said:

This result suggest that although thermal considerations are significant as relates to squish, other mechanical factors (stretching/flexing, etc) are likely to be more significant considerations.

Click to expand...

I did a quick and dirty calc of stretch assuming 32,000 m/s^2 peak piston acceleration, 400 gm piston and average rod cross section of 0.5 inches square for a 6" overall rod lenght and get:


0.011" stretch for a steel rod
0.034" for an aluminum rod

The reality of it is that it is more complicated as the connecting rod is not a uniform cross section.

As for the thermal expansion, I could see some significant differences between a cast iron and aluminum cylinder barrel since the cylinder barrel likely sees the greatest change in temperature.

One builder of race motors would do the initial build and set clearances based on his past experiences and then do an initial dyno pull and then remove the cylinder head to check and "make final adjustments to the clearance". It works but you need to have the knowledge and experience. Calculations can be fun to get a general understanding of what is happening but....................

Dances with Shrapnel said:

WZ507 said:

This result suggest that although thermal considerations are significant as relates to squish, other mechanical factors (stretching/flexing, etc) are likely to be more significant considerations.

Click to expand...

I did a quick and dirty calc of stretch assuming 32,000 m/s^2 peak piston acceleration, 400 gm piston and average rod cross section of 0.5 inches square for a 6" overall rod lenght and get:


0.011" stretch for a steel rod
0.034" for an aluminum rod

The reality of it is that it is more complicated as the connecting rod is not a uniform cross section.

As for the thermal expansion, I could see some significant differences between a cast iron and aluminum cylinder barrel since the cylinder barrel likely sees the greatest change in temperature.

One builder of race motors would do the initial build and set clearances based on his past experiences and then do an initial dyno pull and then remove the cylinder head to check and "make final adjustments to the clearance". It works but you need to have the knowledge and experience. Calculations can be fun to get a general understanding of what is happening but....................

Click to expand...

Interesting figure, would have to add crank flex as well.

Elastic engines are almost biological near limits. I suppose one could put a crank under a press and see its deflection and maybe even better is rig it so the press pulls down on a rod thru the rod ends after heating well and see the total of all the parts in line of load. I want to measure crank sling with a soft witness bolt run up against rim of flywheel then back off some thread count and rev up and creep bolt in till just seeing marks till who know how much matters?

I don't think there is much point for squish bands on a racing Norton. A serious racer will not shift if He's passing someone going into a turn and will over rev the bike. Shifts will also be missed. When that happens the crank will flex because of the wide distance between the main bearings and the pistons will hit the head. I had damage on my pistons with .050" squish clearance. And .060" clearance shows combustion in the squish area making the squish ineffective. And who has ever shown any HP gain on the dyno due to just the very narrow squish band of a Norton? Factory short strokes had no squish band at all with their fully hemispherical heads and they had problems keeping them together because of the HP they put out. When I had to, I revved my racers till the valves floated with racing springs shimmed to 1/32" from coil bind - that was my redline. As bad as I abused my motors - they never blew. But the cases and cranks would crack - needing inspection and replacing (lightweight pistons helped avoid failure). You can avoid over revving and build a mid range motor. But I remember the feeling when passing those bikes on the straights and pulling away. Later - I installed a rev limiter but I ended up setting it so high to avoid my motor cutting out that it was sort of useless.

Lots of top end in my early 750 racer below. Note the homemade carbs. This bike took 2nd place behind my friend Fred Eiker on a Commando at the 1984 750 BOTT Laguna seca National. Fred was a great rider and we swapped leads nearly every lap that race.

My understanding of the squish effect is that it gives more power under the curve while reducing the tendency for detonation, both good things to opt for.
Perhaps an ignition such as the Pazon Power Smart with built in programmable rev limiter would solve the overevving problem and save an engine from any number of catastrophes caused by this. It might not be what the rider wants in the heat of the moment, but bigger picture, who really wants to destroy an engine by overrevving?

Glen

If no squish bands and shorter stroke were good things in twin motors, early Triumph 750s would be better bikes to race than Commando 750s, especially considering the cams in a Triumph are independently adjustable. I've owned a lot of 650 Triumphs in my life, my 850 is much better.

marinatlas said:

Hi Ken, sorry to hi-jack , but when you said cylinder height is 1.552" , what do you mean exactly, just because I am looking for the common lenght , between barrel base flange and cylinder head top .........thanks for any help!

Click to expand...

Actually, my post said 4.552", not 1.552". But, to answer your question, that is the length measurement between the top surface of the cylinder and bottom surface of the base flange.

Ken

Didn't mean to belittle anyone's use of analysis to get a handle on what is happening in the engine. I do the same thing, and I appreciate the folks here who post their calculations and conclusions. I just meant to point out that most engine builders use the values they have found based on their own experience, and for a Commando, they can vary a lot.

Back in the day, Smokey Yunick did a bunch of dyno comparisons on NASCAR engines comparing the results of large to small squish clearance on the same engine build. His conclusion was that once you get to .060" you are loosing power because the squish/quence area is large enough that the amount of unburned fuel-air mixture it contains significantly reduces combustion efficiency. I recall reading somewhere that he said anything up to .050" works well, but that the goal was to keep it as small as possible without contact.

Back in the early '70s, before steel rods and billet crankshafts were a common sight in Commando engines, and the usual bottom end race build for a flat tracker was a standard stroke 750 with race-prepped stock rods and crankshaft and Powermax pistons, that's what I started out road racing with. I was new to Nortons, and built the engines with parts and advice from Axtell. He said to set the squish clearance at .040", so that's what I did. Never had a problem with pistons hitting the head, but I also never revved the engines over 7200 rpm (at least not on purpose), and normally not past 7000. When I later started racing with 920 conversions, I found that .040" was sometimes not enough, particularly when I let the piston clearance get too large. It's interesting to see the comments from other folks with their experiences. Back when we were both racing Nortons, Jim Schmidt used much higher rpm limits that I did, and it's really fun to see his comments about how that affected the amount of squish clearance he needed. Jim Comstock has had experience with a lot of different Commando parts combinations, and I've learned a lot from his observations. He also used to rev his engines way higher than I did when we were in AHRMA races together. Seems like there might be a common thread here.

It sounds to me like you could need a minimum squish clearance anywhere from .035" to .060" in a high performance Commando, depending on parts used and rpm limits adhered to.

We really have hi-jacked the original post, so maybe any additional squish discussions should start with a new post. Then we could go on to boring lengths discussing whether we really need squish combustion chambers on Commando engines.

Ken

Aw shoot motorcyclers are known hijackers, sheeze. The main function of squish is to avoid detonation on are particular CR per octane and timing tolerated. Ideally mixture around the rim stays cool enough by cooler metal nearness [aka: guench] not to blow up from the radiant photon heating till the trapped mixture is jetted into the burning turmoil to add its torque heating of N2 and humidity. If chambers are turbulent enough near TDC and don't linger there long then squish and its quench function not needed. Ricardo is my mentor on this.

Ah , but you can get Supersonic with Squish .

Fast Eddie said:

lcrken said:

It pains me as a scientist and engineer to say this, but theoretical calculations are not too useful for setting minimum squish clearances in Commando engines. Minimum values, i.e. the point at which the piston is about to hit the head, are determined experimentally, usually at some cost in replacement parts.

Ken

Click to expand...

Indeed Ken, experimentation is the only way to be sure with so many (unclear) variables.

In my own experience, Norton crank, alloy rods, iron barrels, racing use, revved to 7500 and occasionally higher; pistons would touch head if gap less than .040, if gap more than .060, evidence of combustion taking place in the squish band area indicated that the squish was starting not to work. So keeping it between .045 and .055 was my own conclusion (with that engine with that usage).

Click to expand...

On my Seeley I have a squish at .055 with full alloy rods (US made, not sure but must be a brand name as RD)
This squish give me satisfaction and no sign of detonation