Valve Seat Concentricity Shocker

[PackardV8]

All I can say here is that a check lap with 800 or 1200 grit paste could go a good way toward detecting excessive eccentricity. DV

Here we go again. There are members here who consider grit paste and lapping to be the devil's work. Do a search to get enough to fill a slow hour.

[WeingartnerRacing]

I have old Sioux with the accu chuck. The machine is amazing. I don’t think they make a new machine that is better. My worse case has been .0005. I attribute that more to the stem being not straight from factory. However you have to have more than .004 before leakage shows up.

I don’t use that guage you use David. An easy test is reface the valve then loosen the collet and turn the valve. Then tighten the collet back up then start grinding. If it doesn’t clean up right away it’s out of round.

I'd be buying a valve run out fixture If I was you. Mine are K Line.

No need as long as it cleans up less than .0005 I am not bothering. I can almost promise a valve is flexing by more than .0005 while running so at some point as person can spend a lot time for nothing.

[MadBill]

Speaking just to torque stress, I read of someone's experience with a heavily-ported Alfa Romeo DOHC aluminum 4 cyl. head (4 stud pattern). Seats and valves were machined with utmost precision and vacuum tested. A plate was then fitted and progressively tightened. At 25 lb-ft. torque, seat leakage developed.

[Schurkey]

I have seen "torque plates" for cylinder heads. I don't know what the difference is between a torque plate for honing cylinders, and a torque plate for valve jobs. Seems to me that a typical honing plate could be bolted to the head instead...but I've never done it.

Now I'm interested in trying.

'Course, that still doesn't run hot water through the casting, or apply combustion pressure...

[statsystems]

I have old Sioux with the accu chuck. The machine is amazing. I don’t think they make a new machine that is better. My worse case has been .0005. I attribute that more to the stem being not straight from factory. However you have to have more than .004 before leakage shows up.

I don’t use that guage you use David. An easy test is reface the valve then loosen the collet and turn the valve. Then tighten the collet back up then start grinding. If it doesn’t clean up right away it’s out of round.

I'd be buying a valve run out fixture If I was you. Mine are K Line.

No need as long as it cleans up less than .0005 I am not bothering. I can almost promise a valve is flexing by more than .0005 while running so at some point as person can spend a lot time for nothing.

^^^^^^^^^^^^This^^^^^^^^^ I've seen valves that were BENT but still passes a vac test. The customer had me vac test them while he was standing there. Pulled up a full gauge. Had a funny look on his face. I say "what's up"? Then he tells me he has Pistons that show marks where the valves were hitting.

6 of 8 intakes were bent. All of them sealed. I guess when you have 360 pounds on the seat you can get forced valve seat conformity.

[swampbuggy]

Statsystems, i would totally agree with you on your last line----up to a certain amount of valve un trueness . Mark H.

[modok]

I agree I have not seen a small amount of runout cause leakage, ever, if the seat is round the valve will flex and sit on it fine.

I was going to point out that NEW heads, after they run a short time, it's not strange to see the seats move .002" or more. So, if a new head has .002-.004 seat runout it would be kinda stupid to grind them all perfect. It would be better to run it for a while and season it and THEN grind them all perfect. OR, if you know which way they will warp.....leave the ones that are allready crooked int he right direction

[ratsnotmice]

I wonder how many cylinder head manufacturers check for valve face to stem concentricity...or even care for that matter? When you receive shipments of 10,000 valves and put together 100 pair a day, quality control is a matter of statistical probability.

[mag2555]

I don't think that running the hottest water that you could thru a Head will get you the same local expansion rates as a running motor getting beaten also up by cylinder pressure and spring forces, but it's likely better then nothing!

[Keith Morganstein]

All I can say here is that a check lap with 800 or 1200 grit paste could go a good way toward detecting excessive eccentricity. DV

Here we go again. There are members here who consider grit paste and lapping to be the devil's work. Do a search to get enough to fill a slow hour.

A few of my past posts on this subject.

I don't like using lapping compound to see the seat contact / concentricity. I feel it's misleading. Lapping compound tends to just show the largest area of seat contact all around the valve because you are twisting when you lap. There is usually enough viscosity to the paste to gray up the whole seat unless it's really bad.Lapping may however repair slight irregularities in the seat. You can lap a valve, have everything look nice and gray, and then it still can fail a Prussian blue check or solvent test.

For checking valve to seat contact, the best bet is to use Prussian blue paste. Smear a thin coating on the valve, tap it with your finger into the seat, then push it out (without twisting) and you will see the contact pattern.

Unless you have some wild guide clearance, the pilot mounted runout gauge will be reasonably accurate.

I fail to see how lapping paste indicates anything in regards to runout. It just transfers the widest and highest spot around the valve face. The paste will exaggerate and this because of it's grit and thickness.

I've posted this before and feel it's the only reliable way to see the contact patch.

Check seat to valve contact with Prussian blue paste. Clean the seat and valve. Apply a thin smear of Prussian blue paste to the valve face.

Position the valve in the guide and drop the valve into the seat. ( A tap with the finger will help it hit the seat) Do this once , with NO rotating of the valve. Pop the valve out pushing from the stem end and carefully remove the valve so as not to smear the paste. This will give you an accurate contact patch.

BTW, you will want to wear gloves with Prussian blue, have a rag handy and be careful you don't make everything you touch blue. Tools, clothes, ears, etc...

[hpetew]

I thought that the biggest worry with a less than perfect concentricity was that that stem would work harden and snap from the excess flexing. I've seen intake valves that were bent to the point that you couldn't re face them but they held compression.

[KnightEngines]

I'm doing a torque plate valve job on Monday.
Big HP Subaru heads, we've had the issue where they are perfect (well, no such thing as perfect - but pretty damn good) when I put them together, screw them down on the motor with 14mm head studs torqued to 135ft/lb & the heads distort enough that the valve clearances move up to .003" & the valves leak straight away.
I'd get them back in for a freshen & I could see the valves were not sealing, but when they were taken off the motor the seats would move back to almost where I cut them & they'd seal up a bunch better.
So I'm gonna have to use a torque plate & stones to finish valve jobs on the damn things when they will be used with the big studs.

[Racerrick]

for collet grinder you have to use a dial indicator mount on your valve grinder to get true valves. First, you chuck the valve just enuogh so when you and turn by hand the collet doesn't rotate. doing this and the machines collet becomes your runout fixture. slowly turn the valve and mark the high spot with a felt, then turn on the machine and watch the dial indicator to see if high spot is in the same felt mark. If it is you can grind the valve if not rechuck it until it is. I know of no other way as I have never seen a collet machine consistently chuck a valve the same way 10 times in a row. That is why I now grind with a centerless grinder. Most of the time valves that come from other shop have .002-.004 runout. very rare to see under .001 unless it is from a centerless grinder.

[modok]

When I run a sioux grinder with the three ball chuck I get average .001 runout, .002 at the worst. If another guy runs it it's usually worse, because they haven't taken the time to examine the issue and find the best procedure to chuck a valve.
I think this is good enough for what I use it for.
There are far more accurate collet chucks for the machines for those that want to see under .001 all the time, but not as versatile at fitting different sizes of stems. Each way has pros and cons as usual

[ProPower engines]

When I run a sioux grinder with the three ball chuck I get average .001 runout, .002 at the worst. If another guy runs it it's usually worse, because they haven't taken the time to examine the issue and find the best procedure to chuck a valve.
I think this is good enough for what I use it for.
There are far more accurate collet chucks for the machines for those that want to see under .001 all the time, but not as versatile at fitting different sizes of stems. Each way has pros and cons as usual

Funny you mention that can you explain in detail what set up procedure you use before grinding the valves.
I have several Sioux machines and they all will hold less then .0005 run out if the valves are all done the same way every time. I learned a couple tricks from the Sioux salesman back in the late 70'2 and later the new guy told me the same routine which is what I still use to this day. But just a hint the chucks are adjustable
Let me know what you system is I may be able to get you to 0

Dave