"Light tight" valve jobs

[joespanova]

In freshening up the 496 in my 69 Chevelle, I'm down to the heads ( last hurdle ). After "cleaning" and reassembly is attempted , I notice the seats look consistent , at least as far as a "visual". The valves after cleaning look "normal" also. These are Brodix Race Rites with 2.250 /1.88 and ovals.
So my usual M.O. is to lap the valves to see how the seats and faces look. One seat ( intake) was bad enough that it wouldn't lap in over about 30% of the seat. On 2 seats ( intakes ) the top angle has a "chipped out" area just above the 45 ( not more than the top angle ) . On those 2 seats the 45 is intact. I'm going to assume the seat distortion on the one seat and the "chipped " areas on the 30 , on the 2 others , was a bad tune up on the previous owners part. Read that detonation. He had a TON of timing in this 10-1 engine and used pump premium. Bracket raced often. Heavy street car.
On all the other seats , they "lapped' in nicely BUT , they are NOT light tight at all. My lapping compound is not super fine so it can cover up a bad seal , I guess. If I apply pressure by finger I can ALMOST get them light tight. Jon Kaase suggested valve jobs don't even have to completely seal to be effective....in fact , he suggested even if I do get a complete light tight seal , the improvement probably wont net any thing. He stated he's seen this over and over again..........BUT , that being said , by nature we STILL want the perfect " static" valve job.
IF, and this is where it gets "dicey" ,if...........they are light tight with the springs you'd think it'd be fine..............BUT?
On my race stuff the valves ALWAYS seal light tight by hand.
ANY THOUGHTS?

[cjperformance]

It is great to have a seat that is concentric to the guide and allows full valve contact (sealing) while not placing any side load on the stem/guide. Often what happens with a torqued down and hot cylinder head is different to what happens on the bench.
At least starting with a near perfect valve seat allows you to asses what may be going on in a running scenario after some run time and an inspection of the seats and valves. Not always practical unfortunately.

[DrillDawg]

See if you can get a vacuum pulled on them with no springs. If you can call it good. Light oil on the seat and hold down the valve with your thumb\finger to get the seal started.

[PackardV8]

Often what happens with a torqued down and hot cylinder head is different to what happens on the bench.
At least starting with a near perfect valve seat allows you to asses what may be going on in a running scenario after some run time and an inspection of the seats and valves. Not always practical unfortunately.

For true. In a running engine, the valve never seats in exactly the same spot twice. Get them perfect on the bench and then install a torque plate; different pattern. Then, just imagine WIGO in a combustion chamber with 500-degree surface temps and 1000-degree exhaust gas heat and 400# spring pressure at 7,000 RPM.

[Keith Morganstein]

In freshening up the 496 in my 69 Chevelle, I'm down to the heads ( last hurdle ). After "cleaning" and reassembly is attempted , I notice the seats look consistent , at least as far as a "visual". The valves after cleaning look "normal" also. These are Brodix Race Rites with 2.250 /1.88 and ovals.
So my usual M.O. is to lap the valves to see how the seats and faces look. One seat ( intake) was bad enough that it wouldn't lap in over about 30% of the seat. On 2 seats ( intakes ) the top angle has a "chipped out" area just above the 45 ( not more than the top angle ) . On those 2 seats the 45 is intact. I'm going to assume the seat distortion on the one seat and the "chipped " areas on the 30 , on the 2 others , was a bad tune up on the previous owners part. Read that detonation. He had a TON of timing in this 10-1 engine and used pump premium. Bracket raced often. Heavy street car.
On all the other seats , they "lapped' in nicely BUT , they are NOT light tight at all. My lapping compound is not super fine so it can cover up a bad seal , I guess. If I apply pressure by finger I can ALMOST get them light tight. Jon Kaase suggested valve jobs don't even have to completely seal to be effective....in fact , he suggested even if I do get a complete light tight seal , the improvement probably wont net any thing. He stated he's seen this over and over again..........BUT , that being said , by nature we STILL want the perfect " static" valve job.
IF, and this is where it gets "dicey" ,if...........they are light tight with the springs you'd think it'd be fine..............BUT?
On my race stuff the valves ALWAYS seal light tight by hand.
ANY THOUGHTS?

Always start with the best valve job that you can. Your heads need freshening up. It’s always going to get worse. Non-concentric seats also lead to valve fatigue and failure.

I remember Kaase’s comments and I don’t believe he was endorsing bad, non concentric, damaged or leaky valve seats. All he was saying was that he hadn’t measured a performance increase on the dyno when going from some leakage to a brand new, perfect as he could make it valve job.

[Zmechanic]

Non-concentric seats also lead to valve fatigue and failure.

+1 to that. At best it will wear the angles out in the seat and ruin the guide. At worst it will snap the valve head off.

[Walter R. Malik]

Light Tight is great but, concentricity is essential.

[looper]

Hi joespanova,
Sounds Like Its Time For A Valve Job. Will The Spring Force The Valve And Seat To Make Good Contact? Probably. But Its Also Forcing The Stem Into The Guide. That Will Cause Guides To Loosen Up Or Valves To Flex.
They Need To Have As Close To Perfect Alignment As Possible When The Head Is Cold And Not Torqued Up. Because Things Will Move Around After. The Closer It Is Now The More The Head Can Distort Before Concentric problems Cause Sealing Problems.
It May Have Had Seats Cut Not Concentric Before You Got Them Or Valves Not Square (Bent).
As Far As Chipping, Not Sure, Its Hard To Hurt Intake Seats. Unless Something Hits It.
Good Luck.

[joespanova]

OK , GUYS , thanks!
All the exhausts are fine , so off I go to the machine shop to touch up all the intakes.
All the exhaust are "light tight" .
Thanks again.

[engineguyBill]

Light Tight is great but, concentricity is essential.
[/quote



X2!! A light-tight seal on the bench is not essential, as the valve will contact the seat at different locations, each time the valve closes, as the engine is running. Concentricity of the valve face to the seat will ensure the the best possible seal. It is also typical to grind the valve with 1 degree difference between the valve and the seat, which helps promote valve seal during operation.

[pamotorman]

the 1 degree difference in seat angle is the answer to a good seal. GM does this on production engines. check the SAE spec sheets

[Schurkey]

My lapping compound is not super fine so it can cover up a bad seal ,

The Permatex crap sold at parts stores is so coarse you might as well just grab a handful of sand from the beach, and mix it with some wheel bearing grease. I bought 400 grit, and I probably should have gone even more fine.

Some folks are against lapping at all. They say the grit embeds in the pores of the metal, can't be fully removed, and causes seat/face wear. I don't know what to think about that.

See if you can get a vacuum pulled on them with no springs. If you can call it good. Light oil on the seat and hold down the valve with your thumb\finger to get the seal started.

I use a vacuum tester. The guy I bought it from (used) said that anything over 50% (15" vacuum) was fine. First set of heads, I was getting 87% (~26 inches) (valve seals installed.) Completely blocked-off, my tester will only generate 89% vacuum.

the 1 degree difference in seat angle is the answer to a good seal. GM does this on production engines. check the SAE spec sheets

I know that this is common, but is it "right"? I thought it was a "cheat" to make the numbers look good at valve-job time, but promoted faster wear.

I do very, very few valve jobs...but I don't cut an interference angle.

[cjperformance]

My lapping compound is not super fine so it can cover up a bad seal ,

The Permatex crap sold at parts stores is so coarse you might as well just grab a handful of sand from the beach, and mix it with some wheel bearing grease. I bought 400 grit, and I probably should have gone even more fine.

Some folks are against lapping at all. They say the grit embeds in the pores of the metal, can't be fully removed, and causes seat/face wear. I don't know what to think about that.

See if you can get a vacuum pulled on them with no springs. If you can call it good. Light oil on the seat and hold down the valve with your thumb\finger to get the seal started.

I use a vacuum tester. The guy I bought it from (used) said that anything over 50% (15" vacuum) was fine. First set of heads, I was getting 87% (~26 inches) (valve seals installed.) Completely blocked-off, my tester will only generate 89% vacuum.

the 1 degree difference in seat angle is the answer to a good seal. GM does this on production engines. check the SAE spec sheets

I know that this is common, but is it "right"? I thought it was a "cheat" to make the numbers look good at valve-job time, but promoted faster wear.

I do very, very few valve jobs...but I don't cut an interference angle.

Sand and bearing grease excellent !

I used to do 46* seats with Neway seat cutters a long tine ago. It was extremely easy to feel if the seat was concentric as the valve would stick on the 46 seat far easier than a 45 seat. Once the seat is true yes they seal very well in quick time so I see why the factory did it. 46 also seemed a little more forgiving to a slightly off seat.
With older style slow ramp cams and soft springs they last fine when done right. That said I think a good 45 seat is better than a good 46 seat.
I dont have a problem with using fine lapping paste for a very light witness lap.

[modok]

so what if you have a customer that goes in the bathroom with your head bare, and turns out the lights? don't panic. Maybe this is ok?
Get prussian blue and use it to read the impression of the seat.
Adjust the angle of your seat or valve so you have even contact across. it does not really matter if it is 44 or 46

You CAN use light to test valve seats, but most will agree it is not the most practical method. There is no direct correlation between this and how it seals or runs. The easiest way to pass the light test is make the angles match exactly, however, as discussed they do not need to match perfectly.

[pamotorman]

so what if you have a customer that goes in the bathroom with your head bare, and turns out the lights? don't panic. Maybe this is ok?
Get prussian blue and use it to read the impression of the seat.
Adjust the angle of your seat or valve so you have even contact across. it does not really matter if it is 44 or 46

You CAN use light to test valve seats, but most will agree it is not the most practical method. There is no direct correlation between this and how it seals or runs. The easiest way to pass the light test is make the angles match exactly, however, as discussed they do not need to match perfectly.

I also used prussian blue paste to check my valve seats. https://www.permatex.com/products/speci ... sian-blue/