Can IC Engine Compression Testing with/without Blanks on Inlet and Exhaust Manifolds Quantify Valve Leakage? 1

[yamoffathoo]

I am considering refurbishing a 200,000 mile 2.4L, 4 cylinder engine sitting in my shop.
Compression testing at 250 rpm yields 90 psig, which increases to 120 psig with 3 squirts of oil in the cylinders, which is below the 140 psig minimum.
To determine whether inlet and exhaust valves are contributing and need lapping, would blanking both manifolds, then each one at a time and recording changes in peak compression numbers yield useful results?
How much compression increase in either valve train would indicate that lapping is required?
Lifter gaps are within specifications, so I am looking for a reason not to disturb them at this time.
Any suggestions would be welcome, thanks.

 

[yamoffathoo]

Hey Lou,

I was pleasantly surprised by the accuracy of this method.

The engine I am assessing for refurbishment is a replacement for one that punched the #4 connecting rod bearing cap through the upper oil pan (casing) at 100,000 miles as the result of a radiator fluid leak. I tore that one down, measured the big end bearing wear on all cylinders, reassembled and verified the measurements using the dial gauge and degree plate (see photo).

The variable that worries me on a high mileage engine is piston ring to cylinder grip, because if it is not there, the gap will not be detectable.

Perhaps measuring with the engine horizontal would reduce that source of error.

 

[BrianPetersen]

Only works in theory if the sparkplug hole is straight up and down aligned with the cylinder bore and in the centre ... which a lot of them are. (4-valve-per-cylinder DOHC ... although some Honda engines are 4-valve SOHC with an inclined sparkplug)

But it still probably won't work, Pistons can rock (tilt) within the bore. There is intentional clearance between the cylinder bore and the piston, and it's significant enough when everything is cold for the piston to measurably rock in the bore. If the sparkplug is not absolutely dead-centre with the wristpin, that'll throw this off. (The intent is that when the engine is running and fully warmed up, the thermal expansion of the piston tightens it up in the bore. They are all like this)

I have the remnants of a knackered single cylinder engine in my workshop somewhere. I knew the big end bearing was stuffed when I looked into the sparkplug hole (inclined to the cylinder) and saw that the carbon had been removed from the area around the perimeter where the piston is supposed to approach the head closely, but not touch. The removed carbon indicated that the piston had been colliding with the head, due to excessive bearing clearance. This was on top of the oil having forbidden sparkles in it when drained.

Don't shortcut this. Follow the normal procedures, they're the way they are for a reason. If you're going to diagnose the engine, follow the proper standard tests, then take it apart and physically measure stuff. You already know it's got low compression. In order to fix that, it has to come apart. With the engine on the bench, it isn't a great deal of work to take the oil pan off and have a look in the bottom end.

You've not told us what engine this is. It could be something that's a dime a dozen at the scrapper's, because there's no demand for used engines, because it's one that never breaks (those usually get squashed and shredded along with the rest of the car because they're next to worthless because there's no demand). It could be something that's worth thousands for a good one, if you can find one, because they ALWAYS break, but at the same time, aren't worth repairing because so much gets trashed when they break that the cost of the parts and labour makes it not worthwhile (*cough* Hyundai/Kia). If it's a worthless one, you've already established that the one in front of you is not healthy (low compression) ... throw it in the garbage and buy a (hopefully) better one from the scrapper's. If it's an expensive one, good luck pricing out parts when it comes to that.

 

[BrianPetersen]

To address a simul-post: It looks like the engine in question indeed has a central vertical sparkplug, in which case, this method can be a screening test. If there's minimal lost-motion of the piston relative to the crank then it probably doesn't have excessive rod-bearing wear. On the other hand, if it DOES show lost-motion, you won't know why. It could be piston-rock (normal). I don't think you need to worry about gravity overcoming the friction from the piston rings.

BUT.

You could have a rod bearing that's worn down to the copper, seconds from failure, and this test won't pick it up.

I don't think it'll pick up main-bearing wear if it's on a single journal, either.

There are some engines known for wearing out the front main due to a combination of the normal tension of the timing belt/chain plus people overtightening accessory drive belts.

Some engines are known for wearing out the thrust bearings, and this test won't pick that up ... although you can do a crank end-play measurement externally in order to assess that.

 

[TugboatEng]

I can tell you're trying to avoid purchasing tools to do this inspection. Do you have a torque wrench so you can remove rod and main caps? If you do, drop the dang pan and look at the bearings. Most of the time bearing wear can be visually assessed.

Valves can be removed with a hammer and socket. Pop a couple out and observe the margin thickness. If the valves are good, buy the spring compressor and re-install them.

 

[mfgenggear]

Okay, Okay, I'll eat some crow on that idea...but here's another one (please see attachment).

This quick measurement can detect excessive connecting rod big end bearing wear without having to disassemble and do the plastigauge routine as follows:

1) mount a 12" dia degree plate on the drive pulley threaded extraction stud holes
2) install a pointer nearby secured to the casing
3) install a brass rod in a steel guiding pipe that threads into the spark plug hole
4) mount a dial gauge on the steel pipe to detect vertical movement of the brass rod sitting on top of a piston.
5) rotate the crankshaft clockwise up to TDC then continue rotating and note the angle when the dial gauge begins to detect downward movement
6) rotate the crankshaft counter clockwise back up the TDC then continue rotating and note the angle when the dial gauge begins to detect downward movement
7) add up the angles and consult the table on the attachment, which will tell you what the big end bearing gap is.

This method relies on the piston rings being able to grab the cylinder and support the piston weight as it loses support from the rod.

For a 12"diameter degree plate, a gap of 0.01mm converts to 2.34/360*pi*12"=0.245", which is very easy to detect as movement on the degree plate perimeter.

Any comments are appreciated, thanks.

Worthless again.

 

[yamoffathoo]

Thanks for the insights Brian,

Both engines are 2AZ-FEXs, from 2007 Toyota Camry Hybrids.

The "dead zone" just after TDC when there was angular motion and no piston movement showing on the dial gauge (accurate to 0.0005") was always repeatable and very symmetric for clockwise and counterclockwise rotation.

Piston rock should probably be detected just after TDC. The tip of the 1/4" diameter brass rod is hemispherical and these 12.5:1 pistons are flat at the point of contact.

 

[yamoffathoo]

The main attraction of this test is that it can be done prior to purchasing a used car, in addition to all the other tests.

 

[BrianPetersen]

Minor digging suggests that a regular used, complete 2AZ-FE goes for $2000 - $3000, but I don't see any Z variants (the hybrids weren't common relative to the regular version).

A rebuild kit looks to be $600 - $1200-ish depending on how sketchy the source is (the cheaper pricing sounds "too cheap to be good") but I think the pistons are different on the Z variant (due to higher compression) and I don't know about the rings. Crank, rods, gaskets, bearing shells should all be the same or interchangeable. If you're going to buy a rebuild kit, make sure you get one that includes the timing chain ... it looks like some of them don't. The higher priced kits include a complete set of valves. Don't forget that there will be some machining costs to refurbish the cylinder bores and possibly the head.

18 year old car with (looks like) 300 000 miles on it ... 100 000 on the original engine and 200 000 on this? Is this a wise investment, or is it time to move on?

 

[BrianPetersen]

The main attraction of this test is that it can be done prior to purchasing a used car, in addition to all the other tests.

The knockety-knocker will be knockety-knocking if there's appreciable big-end clearance. You'll hear it.

Some engines have a reputation for running like that forever. Others will be like a friend's generator at a campground ... heard him start it for the coffee machine ... ran for a few minutes ... tut tut tut tut tap tap tap tap tap tap TAP TAP TAP KNOCK KNOCK KNOCK BANG!

 

[mfgenggear]

The main attraction of this test is that it can be done prior to purchasing a used car, in addition to all the other tests.

What are you doing?
Going in circles. Here, you don't disclose ?
You have to have the right equipment or tooling there is no short cuts.
As Brian said there expensive investment.
Rebuilding an engine takes an experience mechanic. Hire some one to teach you.
A degree wheel is to verify exact timing.
It's an ol hot rodding tip .
An other trick was stick a rag in the spark plug hole hook the starter to a battery with electrical jump starter. And crank it until the rag pops out.
It's near but not exact TDC. Have the valve cover off. Make sure both valves are closed. Stick a screw driver in the plug hole.
And feel the the piston rotated back and forth .
Move the crank bolt with a breaker bar with a cheater bar. Then do the leak down test.