Tech Help: Valve float, not fully closing? (Long)

[Seraphim38]

Context:
• New engine build, 9:1 compression stock bore, stroke
• Stock heads remanufactured and ported by a local machine shop subcontracted by the dyno shop assembling the car
• Stock cams, springs
• New Kenne Bell 2.1L blower, pulleyed for 15-16 psi (Innovators West 10% overdrive/7.5” crank damper, 3.25”upper)
• Robust dual feed fuel system, 60lb injectors, SCT BA3000 MAF
• 500 gentle break in miles of less than 3k rpm, according to Mcleod’s directions (twin disk clutch)

Results:
• Idles smooth, smells like its running rich. No surging, no smoking, no boost spikes. Doesn’t register any OBD2 codes. Exhibits no signs of major failure other than an extremely low dyno graph
• Dyno of 310hp, 380 torque, overboosting to 19psi of boost. 18 degrees of timing, 11.5:1 flat A/F
• Dyno operator says according to Livelink that no timing was being pulled by the knock sensors/ECU. He bumped it up to 23 degrees of timing and got saw hp or torque improvement, and observed no audible detonation
• Shop took the front cover off and re-checked cam timing and it is supposedly right on
• Compression test came out at 120psi for ALL cylinders, with leak down tests ranging from 12% to 32% on cylinder 8


Shops conclusions:
• They can’t explain it and want to send me home with the car as it is. I have already paid over 11 grand in labor for the engine/trans installation.
• The shop thinks that the new bottom end I supplied has bad rings on all cylinders

My conclusions:
• I think that the valves were replaced wrong or the headwork is jacked, as I don’t understand how the compression test could come out all the same but the leak down is so much different
• I believe that if the bottom end, if it were bad, would be smoking badly. The car when I brought it in to them had fragged a cylinder and had 50 psi in one cylinder and 60 psi in another. They called it the Valdez because it blew oil plumes everywhere, and yet still threw down 400lbs of torque at 12 psi of boost—significantly more power at significantly less boost than a completely new motor
• Based upon a conversation with the supplier of the new short block, I believe that the valves are either not fully closing because the stems are too long, or are too short and not fully opening
• It is also possible that the stock valve springs can’t handle the boost and are floating, but that wouldn’t be evidenced in the static compression/leak down tests

I was going to bet the shop that if we re-ringed it and they reassembled it, that we would get the same hp results, and if so they owe me for the labor to have someone else repair/replace the heads. Then it was suggested that we take the motor to a machine shop and disassemble it together with myself, the best Mustang mechanic in Seattle (who should have done the work from the start) and the owner of the shop that did perform the work all present for the tear down and inspection. Unfortunately, it doesn’t seem that there is a local machine shop that knows modular motors well enough to do it.

Anybody have any ideas? Below is the mod list (all parts were NEW when installed unless otherwise noted):

03 Mach 1 aluminum block w/ cobra crank,
ARP rod bolts
ARP head studs
Eagle Rods
Diamond 11cc dished pistons
Original ported B heads from my car, (66k miles) with $2k in machine work to remanufacture them
Original cams, cam springs
New cam gears and cam chains
Milling high volume oil pump
Innovators West 10% overdrive dampener
Thump RRR Racing billet tensioner
Reichart Racing billet idler pulleys
Gates green stripe serpentine belt
Kenne Bell 2.1L chrome supercharger with a factory mounted 3.25 inch pulley
SCT BA3000 MAF
BBK aluminum cold air intake
Accufab SCTB 1694 CFM throttle body
Siemens 60lb injectors (barely used)
Glennsperformance sumped fuel tank
Dual Walbro 255 pumps
Aeromotive billet fuel filter, regulator
Sullivan fuel rails (used)
Even Flow Cooling mod
Stock water pump (that was on my car before the build)
Moroso PCV oil vent/catch can
AC power disconnected
BBK long tube headers and mid pipe (no cats)
Flowmaster cat-back (that was on my car before the build)
Energy suspension poly engine mounts
T-56 transmission with 26 spline input shaft
Mcleod street twin disk
Pro5.0 short shifter
DM Extended Range SpeedCal
UPR billet firewall adjuster and quadrant adjuster
Dynotech aluminum driveline
Innovate wideband
SCT X3 & Pro Racer Package software
Stock rear end gears, axles
New Nitto NT05 315/35/17 tires, with motor break in miles only

 

[trinity_gt]

Compression test came out at 120psi for ALL cylinders, with leak down tests ranging from 12% to 32% on cylinder 8

At the risk of stating the obvious, something ain't right. A low compression might suggest cam timing being way off (though you said they said it was spot on) but the leakdown numbers, if they're correct, suggest a mechanical issue.

If you pressurize each cylinder (at TDC on compression) with 50 psi of air, where do you detect air leaking? Through the oil filler? The exhaust? The intake?

The shop thinks that the new bottom end I supplied has bad rings on all cylinders

Possible. It's possible the bottom end wasn't assembled correctly, clearances are off, gaps are wrong or incorrectly clocked etc. If you do the test I mention above, you can determine if air is getting past the piston rings by checking for air coming out the oil filler cap when the cylinders are pressurized.

It is also possible that the stock valve springs can’t handle the boost and are floating, but that wouldn’t be evidenced in the static compression/leak down tests

I agree. Springs are unlikely. If the air leak is past the valves, maybe you can check the springs on the worst of the cylinders but I think it's going to be something else.

 

[wmburns]

help me understang. Some questions?

Few questions so I can understand.

Pump gas or race gas? What is normal PSI for a compression test? What is the expected leak down range for a new rebuild?

Just wondering. Why the high CR with the high boost? How are you keeping knock/pinging in check?

What is the stock spring rate? What are the valve sizes?

Is it really possible for 15 PSI in the intake manifold to hold the valves open against the spring pressure?

On the valve float question. Don't valve float at high RPM when the springs are not strong enough to close the valves between open/close events?

 

[Dragstr05]

Sounds odd. Picture of the dyno graph could be helpfull. 310hp/380tq on a 4.6 seems to me its fine, and then falling off after a certain point. Just a guess.

120psi compression doesnt really constitute for a dead cylinder, but it does seem low. The shop could also have used a guage that is not reading correctly. It certainly is possible.

It's strange. A flat 11.5 A/F, with 15psi+ in a 4.6 dohc the numbers certainly are low. I'd really like to see the dyno. The whole story simple does not add up (to me).

 

[trinity_gt]

Is it really possible for 15 PSI in the intake manifold to hold the valves open against the spring pressure?

No, not unless the valve springs are not even good to 25-pounds. A 4V intake valve has a head diameter of 1.457-inches for an area of 1.67in^2. At 15psi, this is only 25 pounds pressure. The 4V spring pressure on the seat is supposed to be ~65 pounds so unless the valve spring is basically broken, 15psi isn't going to cause a valve to hold open.

On the valve float question. Don't valve float at high RPM when the springs are not strong enough to close the valves between open/close events?

No. Well, not generally. Valve "float" can be a combination of things including bounce (where the valve bounces off the seat momentarily after closing) and also when the valve spring itself loses authority over the valve hardware due to vibration harmonics. The latter is caused by mechanical inputs to the spring at rates matching the resonant frequency of the valve spring (i.e. at certain RPM points, usually quite high.) When this happens, the spring basically stops being a spring and valve control is lost. This, by the way, is the reason Formula 1 engines did away with mechanical springs and switched to nitrogen gas closure mechanisms to get 18,000+RPM engine speeds. It's also why "beehive" springs are now popular: the changing physical characteristics of the spring along its length "flattens" its tendency to favor any one harmonic frequency.

Regardless, boost pressure and valve "float" are non-issues when one is talking about leakages of 30% (I've read that 2-4% is typical for a fresh motor) and compressions of 120 on a new engine.

 

[bhuff30]

I agree, with leak down percentages like this, you should hear and feel the air going somewhere. A leakdown test needs to be done again and you need to find out where the air is going out. I'm betting on a piss poor valve job. Listen and test to see if you can feel air coming out the intake or exhaust. It may be difficult to identify on the intake side because of the supercharger.

If the rings were the problem, you'd probably see smoking, and it would probably be seeping oil at the seals. However, try to listen to see if the escaping air is going out the valve cover or crank case breathers.

Again, valve float should not be an issue. Under boost, you will have higher pressures inside the cylinder too, so there is no effect felt by the valve.

 

[wmburns]

Thanks trinty_gt. You read my mind with regards to the valve spring rate question. It didn't seam plausible to me and you have done the math to back it up. I just didn’t feel comfortable coming straight out and saying it.

Is the general consensus the leak down is too high and 120 PSI compression is marginal? I agree that a leak down test should also disclose the avenue of the leak (intake, exhaust, rings).

I did read another post regarding poor performance after rebuild. Turned out that the valve stem was too long which didn't allow the valve to fully seat. I am not sure I understand as I would have thought the hydraulic lash adjusters would have compensated. But I guess there are limits…..

If the valve was too short, that wouldn’t affect the leak down results. It would show up at higher power levels as an abnormal drop off in HP.

Isn’t that a little far fetched to think all 8 rings are bad on a new rebuild? IMO, its more likely to be a build error. Incorrect clocking. Rings upside down. Incorrect bore finish. Wrong size pistons (using standard size in an over sized application). In other words a mistake that affects the entire package.

On another note, my Grandmother used to always say to “trust your senses” (she also used to say that common sense isn’t so common). IMO, if you think it is running rich, it probably is.

For all practical purposes this should be a new engine. As such, it should have high compression and low leak down. After a fresh rebuild, anything less seams unacceptable.

 

[Seraphim38]

I am exceptionally grateful for everyone's contribution to solving the mystery that my shop has given up on. Since the beginning, I suspected something with regard to how the valves are setup, though the ideas about piston size and ring configuration I guess shouldn't be ruled out without scientific measurement/empirical confirmation. I am leaving now for the airport to travel for a week, and when I return I am going to deal with this. I may have found a good enough local machine shop to take it to.