valve lapping

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
On "Wide versus Narrow seats". The Engineering is so simple that having learned it as a young-un, I believed everyone knew this: The pressure of the seal must be greater than the pressure of the fluid, for the seal to work. As you have a fixed spring on the valve, say 1lb. and a fixed area of seat - say 0.01 sq. in. the pressure it will seal is 1lb / 0.01sq.in. = 100psi. But if the seat is wider, - say an sealing area of 0.1sq.in. - then the pressure it can seal is 1lb / 0.1sq.in. = 10psi....
In other words, twice the width of seal is half the pressure it can take, or 10 times width of seal is 1/10th the pressure it can take.
I'll let you have a think about your valves and springs and then tell me that "poppet valves have pressure helping them seal"...
OK It is a bit more complex, but your experience of "narrow seats sealing better" is the simple solution. And "proper Engineering" is that simple. Solving problems. Which is what you did to overcome your "crop-circles". (But you do need to use a bit of brain).
Enjoy!
K2
That is not quite correct. You are confusing the area of the seating surface with the area based on the valve diameter. Let's say you have a poppet valve 1.00 inch in diameter at the seat (neglect the width of the seat for a moment. This yields an area of 0.785 sq. in. This is the effective area of the valve. With the valve closed, the force on the valve due to pressure is F= P*A where A= 0.785.
It could be argued whether the effective area is calculated using the inner or outer diameter of the seat. A "perfect" valve and seat contact would suggest the effective area be based on the inner seat diameter. Since nothing is "perfect", the effective area would be based more toward the outer seat diameter.
Of course the width of the seat does affect the unit pressure of the sealing surface, and narrow seats will have higher conformity and thus better sealing potential.
 
Reading this post is most confusing, but definitely thought provoking.
I hope to build a model IC engine one day, so I pay attent to things like valves and valve seats.

Steamchick seems to be saying that the pressure a valve can seal against is linearly proportional to the valve seat area.

Bob T seems to be discussing the force generated on the valve via pressure applied to the top of it?

Steamchick mentions pressure as a function of the valve spring.

I am not sure we are all comparing apples to apples, at least it is not clear to my feeble mind.

We do know that an intake valve with a light spring, such as the non-actuated ones typically found on hit-and-miss engines, will lift off its seat on the intake stroke, and be sealed to its seat upon the compression stroke.
The force on the seat, and the pressure from below that the valve can seal against, is related to the pressure being exerted on the top of the valve, regardess of valve seat diameter, ie: it is an upper pressure vs lower pressure thing, or the differential pressure between the internal cylinder pressure and atmospheric pressure.

The valve spring is intended to assist the valve to the seated position, not generate the force required to seal the valve (for an IC engine).
For a safety valve on an air tank, the converse is true; the spring on the safety valve determines the pressure at which the safety valve will open.

The exhaust valve springs on hit-and-miss engines are many times stronger than the intake spring, and the exhaust valve is forced open via the cam/pushrod/rocker.
The force exerted on the exhaust valve by the spring is for the purpose of quickly closing the valve, and preventing any valve bounce (I think).
Again, the exhaust valve spring is not sealing the valve on the seat nearly as much as the pressure from above the valve.

No doubt if there were no pressure above the valve, then the spring force would dictate how much pressure the valve could retain, which I think is what steamchick is saying.
But the dictating forces in an IC engine are from internal cylinder pressures; valve springs just act to close the valves after they open.

This is all purely hypothetical, and not based on anything other than what just popped out of my head, but this would seem to be logical.

.
 
I have seen people seat a ball valve on the seat using the method mentioned, which was tapping it with a hammer.
I don't have any experience with IC valves, other than lapping lawnmower valves that get pitted and burned.

I think the key for the "tapping with hammer" thing to work is that the valve and seat are very close to being exactly round and concentric.
A valve or seat that is not concentric would only seal in the position where it was tapped with the hammer, and if/when the valve rotated, it would no longer seal.

When I have lapped lawnmower valves, they tend to become oblong I guess due to heat/distortion, and you can see where the compound begins to cut on about 1/2 the valve surface before gradually spreading to the entire valve surface as the lapping process progresses.

Lapping using the method that Brian Rupnow mentioned is the way I have always done it, using progressively less coarse compound grit.
If the valve and seat (lawnmower) has only light pitting, I start with 600 grit compound.

It may also be true that a light hammer tap will align the valve with the seat.
I have not built any IC engines yet, so I have much to learn and verify.

.
 
That is not quite correct. You are confusing the area of the seating surface with the area based on the valve diameter. Let's say you have a poppet valve 1.00 inch in diameter at the seat (neglect the width of the seat for a moment. This yields an area of 0.785 sq. in. This is the effective area of the valve. With the valve closed, the force on the valve due to pressure is F= P*A where A= 0.785.
It could be argued whether the effective area is calculated using the inner or outer diameter of the seat. A "perfect" valve and seat contact would suggest the effective area be based on the inner seat diameter. Since nothing is "perfect", the effective area would be based more toward the outer seat diameter.
Of course the width of the seat does affect the unit pressure of the sealing surface (This is what I am talking about - K2) and narrow seats will have higher conformity and thus better sealing potential. - Correct.
And here's where I disagree, being pretty sure I am not confused. The force applied to the sealing surface, divided by the Area of the contact face sealing surface becomes the "EFFECTIVE PRESSURE WITHIN THE SEALING SURFACE". This must exceed the fluid pressure - otherwise molecules of fluid will pass the seal. Initially, the pressure on a poppet valve is only derived from the spring pressure, but this does increase as the fluid pressure on the "whole effective area" acts on the valve, as you explain, but unless the initial seal is formed by the spring force and valve-face to seat-face geometry, the valve will leak. Having designed equipment that had to be "hermetically sealed" and having re-designed elastomeric seals that initially proved inadequate, I had to "learn the rules" - professionally.
I apologise if my explanation was mis-leading. I was trying to explain why larger width of lapped surface of the valve seats always leads to leakage of the valve. Oft misunderstood by seemingly "clever mechanics" who think a wider seat (lapped surface) will seal better (I have met many!). There is also a correct comment about conical misalignment - which is a root cause of leakage when valve stems and valve guides wear.
As you explained
Green twin is correct : "
"the key for the "tapping with hammer" thing to work is that the valve and seat are very close to being exactly round and concentric.
A valve or seat that is not concentric would only seal in the position where it was tapped with the hammer, and if/when the valve rotated, it would no longer seal. " - which is why I do not advocate "hitting" poppet valves in any circumstances. Any mis-alignment is likely to lead to distortion of the relationship of valve head to stem alignment. This means that when the valve turns (as it will) the valve will then naturally become mis-aligned and of course will then leak.
Of course, there are always 2 limits in design, Max and Min. We have considered "minimum seat pressure" - ergo "maximum seat area" but the materials and temperatures involved also affect the "minimum seat area", as a seat that is "too narrow" (in the case of both the poppet valve and the pressure vessel safety relief valve) will naturally deform as the yield point of the "weakest" material is exceeded, to become large enough to withstand the maximum forces on the sealing surface (forces derived from the closure spring plus pressure differential upon the valve head). (This is clearly seen when a valve acts upon an elastomeric seal in compression - the elastomer can take up a set due to material creepage with time and pressure. I saw this on equipment after many years of service where elastomeric face-seals were well distorted to the point of leaking, as a result of compression set. But that was a complex problem that needed re-designing.).
Enough said. (Most models are made, and happen to work without the engineering needed for "sustainable and profitable industry").
K2
 
I got the valve cages made, the ends are at 90 degrees..
20211119_154548.jpg


I then put a valve in a cage, held the end of the valve in a collet, then by hand held the cage and spun it against the valve back and forth a few times, lifted cage away from valve, rotated, then repeated a few times. I didn't put any lapping compound on it but it did make a small bevel
20211119_152916.jpg


If I tested valve this way with a mighty-vac, what amount of time for it to go from 30inHg to no vacuum is decent?
20211119_153919.jpg


I made a way to test it with putting air pressure through spark plug, but that was with cages I had build before, and they were loctited in the head already. I want to test them before they get loctited. That's why I figured I'd test it with the mighty-vac
20211119_161228.jpg
 
I think testing it with the mighty-vac this way won't work, because if I pull the valve away from the cage, it still very slowly loses vacuum, because the small hole the valve slides in provides somewhat of a seal.
 
Pressure decay takes a lot of calibration to get right. (from what I learned in industry). The simplest procedure (not "best?", just something you can maybe do "at home?) is a simple calibrator: Using a needle valve (fuel valve from an Aircraft model shop, that normally fits into a 2-stroke aero engine intake - or something?). Take the valve "fully closed" and monitor pressure decay from pump "ON" => vacuum valve closed time to plus 30 seconds - or 5 mins - or whatever interval suits so you get (say) 20% vacuum pressure drop: Then repeat the check, just cracking the needle valve to admit air so instead you get (say) 40~50% vacuum pressure drop. Then you can say "you have a leaker!". That gives you 2 conditions that define an OK and No Good condition for your testing. - I was taught by the guy that commisioned many Helium-leak-test stations for various car parts, and pressure-decay testers.
Pressure decay wasn't perfect though, but fast and effective for production, where less than 6 failures per million parts were allowed. But it needed daily/shiftly calibration against the OK and NG parts, "to be sure to be sure". (The parts held after 1 calibration until the next calibration was confirmed OK - it is an OK "end" calibration check that proves the batch OK for shipping, not the "start" calibration check.).
OR: Assemble the valve into the valve seat body, held by the spring, and introduce a drop of petrol, or other thin fuel, into the valve body so you can see if any fluid leaks past the valve. Just don't smoke while doing this one! You can assemble parts, then black with a candle flame before testing with fuel. The carbon dust shows-up the fuel leak a bit easier than "shiny metal"... Talc may work as well... But extinguish the candle before opening the fuel can! An "eye dropper" or pipette can aid putting a drop of fuel into the valve body...
It works well for car engine poppet valves...
I have lots of "stupid ideas"... just tell me what works and what doesn't!
K2
 
Simple trick to check valve and seat that I learned from my father a long time ago (I learned a lot of knowledge about engines, mechanics.. from my father, for me he is a master)
Make sure valve and seat are clean. .Hold the valve in the seat by hand or spring, pour a little kerosene or similar liquid or water into the IN or EX, . After about 10 seconds, look at the valve and seat contacts (Or use paper towels to wipe them) : if they are dry that's good enough
Or another way: put a little kerosene on the contact, and use compressed air about 1 bar to blow into the IN and EX, if there are no air bubbles in the contact, the valve and seat are fine.
I usually check the valve and seat this way and they are always fine
I did it both ways: Lapping valve with seat, and closed the valve to the seat with hammer
But for me, the Lapping valve with seat is always the first choice !
 
And here's where I disagree, being pretty sure I am not confused. The force applied to the sealing surface, divided by the Area of the contact face sealing surface becomes the "EFFECTIVE PRESSURE WITHIN THE SEALING SURFACE". This must exceed the fluid pressure - otherwise molecules of fluid will pass the seal. Initially, the pressure on a poppet valve is only derived from the spring pressure, but this does increase as the fluid pressure on the "whole effective area" acts on the valve, as you explain, but unless the initial seal is formed by the spring force and valve-face to seat-face geometry, the valve will leak. Having designed equipment that had to be "hermetically sealed" and having re-designed elastomeric seals that initially proved inadequate, I had to "learn the rules" - professionally.
I apologise if my explanation was mis-leading. I was trying to explain why larger width of lapped surface of the valve seats always leads to leakage of the valve. Oft misunderstood by seemingly "clever mechanics" who think a wider seat (lapped surface) will seal better (I have met many!). There is also a correct comment about conical misalignment - which is a root cause of leakage when valve stems and valve guides wear.
As you explained
Green twin is correct : "
"the key for the "tapping with hammer" thing to work is that the valve and seat are very close to being exactly round and concentric.
A valve or seat that is not concentric would only seal in the position where it was tapped with the hammer, and if/when the valve rotated, it would no longer seal. " - which is why I do not advocate "hitting" poppet valves in any circumstances. Any mis-alignment is likely to lead to distortion of the relationship of valve head to stem alignment. This means that when the valve turns (as it will) the valve will then naturally become mis-aligned and of course will then leak.
Of course, there are always 2 limits in design, Max and Min. We have considered "minimum seat pressure" - ergo "maximum seat area" but the materials and temperatures involved also affect the "minimum seat area", as a seat that is "too narrow" (in the case of both the poppet valve and the pressure vessel safety relief valve) will naturally deform as the yield point of the "weakest" material is exceeded, to become large enough to withstand the maximum forces on the sealing surface (forces derived from the closure spring plus pressure differential upon the valve head). (This is clearly seen when a valve acts upon an elastomeric seal in compression - the elastomer can take up a set due to material creepage with time and pressure. I saw this on equipment after many years of service where elastomeric face-seals were well distorted to the point of leaking, as a result of compression set. But that was a complex problem that needed re-designing.).
Enough said. (Most models are made, and happen to work without the engineering needed for "sustainable and profitable industry").
K2
Thank you for the response. We are in agreement that a narrower seat will tend to seal better, but not for the same reasons. Given all else the same, if the form of the valve & seat contact surface are 'perfect' ie zero geometry deviation, 100% contact, smooth surface, etc., a narrow and wide seat width will both seal equally. Since nothing is 'perfect', the narrower seat will tend to have a smaller leak area, plus the higher unit loading will provide additional conformity.
I mean no disrespect on your many years of practical experience. I think we will just have to agree to disagree.
 
I use tooth paste as a lapping compound. Do not know how it compares to manufactered lapping compounds.
I clean the tooth paste off completely whenthe lapping is done.
mike
I was using tooth past too for lapping pistons and liners,crazy amount of work,but great finish and not harmful after even little cleaning...
 
with valve closed against valve cage, around 5 min..
with valve pulled away from cage around 2.5 min..

The rate at which it leaks seems to be more at the start (more vacuum), and slows down as it goes towards normal air pressure
 
so to get the loctited cages that are in the head now (green loctite), I have a heat gun and a propane torch, which to use to heat it up before I try and press the old cages out?

After I get new cages installed, I can perform pressure test by putting air through spark plug hole and listen for leaks, any thoughts on how much psi I should put to it?
 
Would this stuff be good to use for lapping and is 320 grit ok?
View attachment 131154
300 seems a little coarse, but as I mentioned, I have not built a model IC engine yet.
I have only done valve work on lawnmower engines.
I guess the proof is in the pudding though, and if you tried the 320 grit, and it worked well, then no need to find 600 grit.

Brian Rupnow had problems with his valves seating until he got some 600 grit lapping compound.

Does anyone have any data on the life of narrow vs wide valve seats?
Seems like a narrow valve seat would be prone to wear, since the force would be concentrated in a much smaller area on the seat?
Maybe it is a non-issue on a model IC engine.

There must be a sweet spot for valve seat size.
I guess the height of the top of the valve dictates the seat size?
A taller valve top could have a longer face at 45 degrees or whatever angle is used.

.
 
so to get the loctited cages that are in the head now (green loctite), I have a heat gun and a propane torch, which to use to heat it up before I try and press the old cages out?

After I get new cages installed, I can perform pressure test by putting air through spark plug hole and listen for leaks, any thoughts on how much psi I should put to it?

Scooby,

I recently removed Loctited valve cages from an aluminum cylinder head by making a punch that fit the ID of the cage. I heated the end of the punch to a dull red, then used the punch to transfer heat to the cage and press the guide out.

As to pressure testing your valves, 50 psi should be good. A cheap stethoscope with the sensor removed is handy for finding leaks.

Chuck
 
Here's a little more to add to the mix. In automobiles, the only way valves are able to get rid of some of the heat of combustion is direct contact thermal transfer. This essentially means that automobile engines need contact over a wider area between valves and the valve seats to shed heat. For our small engines heat is not nearly as large a problem, so an almost "line contact" between the valve and the valve seat is acceptable.
 
I made this to press it out, pin will keep things centered, and the ID slips over top of valve cage and od is small enough to go through hole in the head
20211120_194742_resized.jpg

20211120_194752_resized.jpg

I was gonna use the lathe tailstock to press it out
 
Scooby, You need to chemically degrade the loctite to relase the parts. Use a bit of a bar of household soap to mark some lines on the aluminium cylinder head,: these will tur black when the cylinder head is hot enough - and to avoid overheating anything - and the loctite should have been adequately burned to non-bonding chemicals. Either hot air gun or flames are OK.... differential expansion (of course) will affect the fit of parts so a press to dismantle is safer than a hammer! (I use my 4" engineers' vice, as it is a screw-press in application). But make suitable supports for cylinder head and pins for the valve seats.
K2
 

Latest posts

Back
Top