A new attempt at making piston rings

[KenC]

I have made dozens of rings from 3/4" - 1-1/4" adhering closely to the Trimble method published in Strictly IC. They work great and I don't think I have ever broken one. I parted them using a ThinBit carbide grooving tool. No toolpost grinder here.

Me too. I have made many rings sticking very closely to the Trimble method with no problems at all. I thought that it had already become the standard way of making rings. I use a 1/16" wide parting tool ground at a slight angle to part close to the right, and I break the rings using a purpose-built cleaver.

 

[cds4byu]

Carl---I'm in Ontario, Canada. Where are you?--Brian

I'm in Utah, USA

 

[Stefan-K]

the piston manufacturers work to 0.0002 ????

This is 2 ten thousands of a millimeter! Are you sure about this ?? Every small temperature change, which you have several times a day will change the dimension of this piston by far more. No one in the Massproduction Industrie will manufakture to such small tolerances.

 

[Brian Rupnow]

Time has come to cry FAIL!!!! I can not get the engine to fire, even with new piston, purchased new cast iron rings and new cast iron cylinder. Valve timing has been reset, ignition timing has been reset, carburetor has been adjusted to every possible configuration, and the engine doesn't fire. Compression feels cheesy when flywheel is turned by hand. If I squirt a bit of oil down the sparkplug hole, the engine has great compression and starts right up and runs like a trooper until the oil on top of the cylinder has burned away. Then it quits. This is purely the result of too low or no compression due to leaky piston rings and "fit" of piston in cylinder. I have monkeyed with this engine (which ran just great with a Viton o-ring) until I am sick of it. I have received a lot of interesting opinions on ring making, piston making, and state of the world in general from forum members and yes, I do read and pay heed to all of it, thank you. Later in the summer, when I have had time to digest all the new information and I'm not so burned out, I will use my new heat treat oven and toolpost grinder and make a new set of rings 0.038" wide and possibly another new piston with grooves to accept the new rings, I may try again. For now though, I'm whipped. Thank you to all of you who have followed this thread, and for the information given.----Brian

 

[Gordon]

You have said a couple of times that your information says ring thickness should be about the same as ring width. According to Mr Trimble who some feel has the ONLY answer the width should be 1/25 to 1/30 of bore. With 1" bore that would be .040 to .033. He says that thickness should be .043 to .045 x bore which would be .043 thick. Using your .038 width for thickness of .038 according to the graph with Trimble that would make the wall pressure below 30 PSI which would give low compression. Next time perhaps try making the rings thicker.

Lots of information out there on model rings. Read 10 articles and get 30 different opinions.

Now put it on the shelf and enjoy summer.

 

[Tim Wescott]

Later in the summer, when I have had time to digest all the new information and I'm not so burned out, I will use my new heat treat oven and toolpost grinder and make a new set of rings 0.038" wide and possibly another new piston with grooves to accept the new rings, I may try again.

+1! The three ways that, as an engineer, I've learned how to get through a brick wall are: (1) find a door, (2) explosives, (3) just beat your head against it until it gives up.

That last one sometimes requires that you take a break.

 

[DKGrimm]

[QUOTE=". . . For now though, I'm whipped. Thank you to all of you who have followed this thread, and for the information given.----Brian
[/QUOTE]
I am where you are with an Atkinson differential engine. Three sets of cylinder, pistons, and rings, and no compression, no explanation. That's why I have been following this thread. My plan has been 1) give it a rest and do something else for a while, and 2) come back and fix it later. All this conversation actually helps.

 

[Brian Rupnow]

This is a good time to ask a theoretical question. If I had a Viton ring on a piston, and shoved the piston half was down a length of cylinder, there would be no leakage at all past the Viton ring. A Viton ring seals perfectly around it's full circumference. If I stood the cylinder on end and filled the top half with Varsol or any other fluid, none of it would leak past the Viton ring---ever. That is a given. I have done that.---Now, if I take a piston with two cast iron rings on it, each ring having a 0.004" gap between the ends, and the gaps set at 180 degrees to each other, it will leak. If placed half way down a cylinder and stood on end, and if I fill the top half of the cylinder up with Varsol or some similar fluid and left overnight, some of the Varsol would leak past the rings and be puddled in the bottom half of the cylinder. Iron rings are not a perfect seal. The only reason that they work in an engine is that during a typical engine cycle, there isn't enough time for the compression to leak down past the rings. Some does--but there isn't enough time for the compression to degrade below a point where ignition is no longer possible. I ask this, because I would like to make a testing rig which would let me know if a piston and rings were "good enough" to make an engine run.-----------Why would I do that? ---Because on a real engine there is always a lot more involved in changing the rings and piston than just the cylinder bolts. All I can think of is a leak-down pressure test. Problem is, the volume of a 1" bore x 1" stroke engine is so small, it would require far better and more sensitive equipment to do a leak down pressure test than anything I currently have.---And even a leak down pressure test is only comparative. I would have to perform the test on a "good" piston and ring set to form a baseline from which I could measure the effectiveness of my own piston and rings. The ONLY way I know how to test the effectiveness of a ring and piston set, it to put it in an engine and see if the engine runs.---Any comments???

 

[mayhugh1]

It's normal for oil to leak past cast iron rings in an upside-down cylinder over an extended period of time. Checking a radial engine for hydrolock before starting is is s.o.p. in both model and full-size engines. I even remove the lower plugs during storage on my radials. These leaks occur through the ring gap and also between the rings and cylinder walls because without combustion, the ring's spring pressure alone isn't enough for a perfect seal.

I was confused about your last attempt. I may be wrong, but your posts seem to say you were lapping your new piston to your new cylinder and also adding store-bought rings that had received some lapping of their own in another setup. If this is all true, I don't think the rings were trustworthy because your previous lapping changed their o.d. if not their circularity. The piston should be a loose sliding in the cylinder and not a close-lapped fit. There may have been too much friction for the little engine to overcome under those conditions.

The only thing thing I can (re)suggest for a fixture to predict a ring's performance is a good light leak fixture like I've described earlier.

I was also curious about your lapping technique. I hold my cylinders in my hand and power the lap with a battery powered drill in the other hand. Lapping is intended to be a slow and easy process. I never try to take off as much at a time as you described and for me, removing a total of .002" from a bored cylinder is near the limit of my patience at the metal removal rate I use. I mentioned earlier that I lap all my cylinders as a group and only a tiny bit at a time in order for all the cylinders to arrive at the same diameter. Sometimes I overshoot a particular cylinder because I misjudged the amount I turned the screw on the lap and the lap jammed in the cylinder creating a scratch that I have to work out. Sometimes this means increasing the size of all the other cylinders as well. Are you sure you didn't damage your cylinder when you (over)powered that .002"(over) lap into the cylinder. Also, I'm not convinced your measurements are as good as you think. We don't have traceable measurement standards in our home shops, but a micrometer over a dial bore gage so you're depending on repeatability rather absolute numbers is almost as good. - Terry

 

[petertha]

Using your .038 width for thickness of .038 according to the graph with Trimble that would make the wall pressure below 30 PSI which would give low compression. Next time perhaps try making the rings thicker.

The Trimble article says 'The first characteristic to consider is is cylinder wall pressure. When installed and operating, the ring must exert sufficient pressure against the cylinder wall by virtue of its own springiness to resist being sucked away from the wall by vacuum in the combustion chamber. It is not unusual for a 4-cycle engine to develop 10 psi suction during during intake strokes under low throttle, high rpm throttle conditions. Using a safety factor of three to allow for our ignorance with respect to uniformity of pressure exerted by a less than perfect ring in a less than perfect cylinder, we can deduce that an acceptable ring should produce at least 3x10=30 psi wall pressure'.

Maybe I've misunderstood this issue all along, but I always assumed the Trimble 30 psi constraint curve was more about low pressure induction stroke, not high pressure power stroke. The was based on notion that combustion pressure was also acting behind the annular gap of ring & piston groove so if anything would assist sealing pressure, additive to ring wall pressure (but still no match for bad / open ring geometry). Attached screen grab from one of Terry.M's posts, possibly the 18-cyl radial. Maybe he can comment. And excerpt from PDF document leerkracht provided.

I've actually lost track of what Brian's ring dimensions ended up being. And of course trying to position his ring on the Trimble chart its complicated with some unknowns; actual bore diameter, heat set dowel diameter (equivalent) & several other factors. But anyways FWIW, here is a spreadsheet version of the Trimble calcs and corresponding plot of constraint curves. I've been re-writing it lately so if anyone spots any errors, please let me know by PM. Trimble's article referenced his own 1"bore V8 which I believe is the same bore as this subject project. So actually, no calculations were even required.

For consistency, Trimble used the word thickness (T) to mean the radial ring dimension viewing piston from top. And width (W) to be the axial dimension viewing ring from side. I've inserted the recommended T/B and D/B target & range notes from his article in the spreadsheet for quick reference, but obviously the article provides more complete basis for these values.

 

[petertha]

Time for Ringmaking-Take2. I am now going to make 4 more rings using the Trimble? method. I do have a 0.020" slitting saw. So---the rings will be turned to (3.142+.020)-3.142=1.006" o.d. and 0.898" inner diameter. This will give an inner diameter the same as rings I have purchased form Debolt. They will be cut to 0.063" thick. Instead of being broken, these rings will be cut using my 0.020" slitting saw. No farther work on cleaning up the i.d. nor the o.d. nor the thickness, until I have spread them individually over a 3/16" piece of steel inserted into the sawn gap and heat treated them to cherry red and then allowed them to air cool. Then they will be polished to 0.062" wide and cleaned on a sheet of emery paper on a flat surface, then positioned in the cylinder bore and if needed the gap will be filed to give a 0.004" wide gap.

I know this is going back a ways (post #57) and please don't take this criticism of what methods & effort you eventually employed. But the Trimble? (question mark) is appropriate because its just not the Trimble method on multiple fronts. I'm not saying what's right or wrong, just reading the SIC article and making the comparison.

- nowhere does Trimble mention a slitting saw. He goes into quite a bit of detail saying consistent success eluded him until he eventually settled on cleaving the rings. Basically trying to get as close to a zero width part such that the rest of the procedure & heat setting apparatus would ensue with a known reference. The only gap treatment prior to heat setting was minor clean. Basically eliminating a micro carrot of material stuck on one side of ring which originated from corresponding carrot hole on other side. The protrusion had to be levelled so the two faces aligned properly to the dowel diameter as intended.

- other folks have offered input as to how to compensate the OD if a slitting saw is uses. No comment because of many examples where it obviously works. I'm just saying it wasn't in Trimble's method. The post heat treat gap width calculation is defined in the writeup.

- the pre-heat set Trimble ring OD & ID are fully defined in his description. The OD has a very minor +tolerance but he says toss any rings that come out under bore B. The ID is a calculation result of the T/B ratio, not related to the piston or anything else. ID = OD - 2*T

- re heat set, there is an interesting chart in the article (Fig-12). It took me a while to digest it & anyone please comment if I'm misinterpreting. He is showing the effect of supporting the open ring on a rectangular spacer block as opposed to on the dowel. However the gap opening between the two setups are identical. The difference is that the ring is supported on its inner corners in the case of rectangle plate vs. resting on the ring neutral axis tangent point in case of dowel spacer. The resultant plot then shows the X-axis around the bore and Y-axis delta R (radius in inches) deviation to the bore. The rectangle spread curve deviates as much as 0.0035" from bore wall vs. the dowel based curve which he is saying is the theoretical perfectly circular curve plot (consistent radial seal). 'the dotted line shows the same data for the ring spread with a spacer as in fig 11, it has the proper curvature in two places where the dotted and solid lines cross, Everywhere else the curvature is too large or too small. Particularly disturbing is the area near the gap where the curvature is less than the cylinder curvature, There is no way for this area of the ring to touch the cylinder when installed because it is more sharply curved than the cylinder itself. So this seemingly minor subtlety with the same effective opening results in basically a barn door sized opening relative to high pressure gas. The significance of varying the dowel diameter would be greater still, which is why he provides the T/B vs D/B constraint chart.

Also note, the calculated dowel diameter is not actually the maximum opening for ring during heat set. I had to stare at those unfortunately teeny SIC diagrams & re-read the verbiage to realize - its the combined effect of his mandrel diameter, dowel diameter & their relative spacing on the fixture that positions the ring with the correct opening with dowel contact at the neutral axis as described above. And this fixture is intended to assist so its hard (er) to misalign ring in spread mode.

 

[minh-thanh]

You use oil in the combustion chamber to test compression, and you have compression (I only use a little oil at the rings - very little -), so
Check everything: is it completely airtight?
Valve ? spark plugs? cylinder head...
Because can leak and with oil it gets better
With the CI ring and piston there is a bit of a leak so things get worse: too little compression,
check the cylinder and rings with light: the cylinder and ring must be dry and clean, if they have some oil you won't be able to know if it's good enough

 

[Tom Watson]

during my motor cycle racing days when i was young and stupid, I used Dykes rings. See attached diagram. (a am poor artist). They were a fairly low pressure on the bore, but the combustion pressure caused a much higher pressure on the back of the ring. This gave a lowering of the ring friction of the ring on the bore but the compression was excellent. They only gave good compression while the engine was running not when pulled over very slowly.

when turned over slow the compression was poor. The ring was located level with the top of the piston.
I don't know how they would work on our small diameter rings. I should try one.

 

[johwen]

HI Brian,
As a last ditch effort. Why not follow the Trimble method to a T, and get a ring that works. From your descriptions you seem to dodge around the full method describe by Trimble. The heat treatment as described is an important part of of the construction particularly the spread of the gap too much and you end up with a ring out of round and too little and you end up with too little wall pressure. BUT MOST IMPORTANTLY DON'T GIVE UP as the Trimble method is proven and works. John

 

[jack620]

the piston manufacturers work to 0.0002 mm when machining

I seriously doubt that.

 

[Gordon]

The significance of varying the dowel diameter would be greater still, which is why he provides the T/B vs D/B constraint chart.

Also note, the calculated dowel diameter is not actually the maximum opening for ring during heat set. I had to stare at those unfortunately teeny SIC diagrams & re-read the verbiage to realize - its the combined effect of his mandrel diameter, dowel diameter & their relative spacing on the fixture that positions the ring with the correct opening with dowel contact at the neutral axis as described above. And this fixture is intended to assist so its hard (er) to misalign ring in spread mode.

Most of Trimble math and calculations are beyond my ability to understand them but I am unsure if a slight deviation in dowel size is a problem as long as the mandrel diameter and the offset are adjusted accordingly. With the small dowel as calculated it is difficult to get a pin sized to the correct diameter. For instance in his 1" bore example the dowel is .150. 5/32 is .156 and is a standard rod size. Would using the oversized dowel be a problem as long as the mandrel and offset are adjusted accordingly? A 1 1/8 bore would give a dowel of .16875. Would a 5/32 diameter be too much variation?

Gordon

 

[kuhncw]

Gordon, you can make the end of the dowel a common drill size to fit the hole in the flange of the mandrel and the turn the part of the dowel that contacts the rings to the calculated size. That has been my solution.

Chuck

 

[Tim Wescott]

Hey Brian:

I'm not sure if this is useful, or if there's a way to make a formal test out of this, but with a model airplane engine I rotate the prop through a compression stroke and feel how long it holds compression. It's all touchy-feely, but it's worked well for me.

I'm just going to tell you what I do, and if it's helpful you can translate this into a fixture:

For a model airplane engine, if you've got compression when you give it a good hard flip it'll start and run at full speed. If you can rotate it through TDC at about one rotation per second without palpable loss of compression then it'll probably idle reliably. All the engines I've owned have been lapped -- a lapped engine in really good shape will retain palpable compression at TDC for at least a second, but if it's doing that well it's just gravy.

Another good way to assess piston/cylinder seal on a model two-stroke is to make sure there's some fuel in the cylinder, then rotate the thing through TDC while watching the piston through the exhaust port (so -- definitely a two stroke thing). You'll always see some bubbling from air leakage, but less is better and more is worse.

 

[Tim Wescott]

Could you make a small-bore pressure gauge that just directly screws into your spark plug hole?

Or cheat: hie thee to the nearest auto store, buy an el-cheapo tire gauge and replace the business end with something that screws into your favorite size spark plug hole. I'm not sure of the details of what's inside of those, but it's got to be lower displacement than your 1" bore engine.

(And I'm suggesting a tire gauge on purpose because they capture the maximum reading automagically, with all the mysterious stuff that's inside them).

((Now I have to look and see how one is made)).

Even with some error due to the gauge's displacement, you should still get a comparative pressure reading.

 

[DKGrimm]

Could you make a small-bore pressure gauge that just directly screws into your spark plug hole?

I've had good luck with this homemade leak-down compression tester. I put a regulated 40 psi on the input and measure the cylinder pressure with the gauge. There's only a tiny hole for the air to get to the cylinder, so a compression leak will drop the cylinder pressure. Another advantage is that you can listen to see if the leak is coming through a valve or through the crankcase via the rings.

Don