Showing posts with label leakdown. Show all posts
Showing posts with label leakdown. Show all posts

Sunday, November 24, 2019

#ChelseaTheCivic, Part 5-- Off-Season Engine Health Assessment

Introduction

Racing puts a lot of stress on a car.  In addition to the lateral and longitudinal Gs subjecting the chassis, structural suspension components, and bushings to stress, cars get crashed, transmissions break, driveline components break, and shocks wear out.  Oh yeah, and engines get abused too, I guess.

Ant working on Alex's #TrashTeg

For example:

After living in a 1998 Acura Integra for 170,000 miles, my B18B1 was transplanted into a 1993 Honda Civic DX for racing duty.  

Denise the D Series Integra, Donor to Chelsea the Civic

While in the Civic, it experienced a couple of money shifts, I found some metal shavings in the oil a couple of times (not sure from where), I lost a radiator cap and pegged my coolant temperature in a qualifying session, and toward the end of this season, it began smoking more than usual, and halfway through the season, I also did a fuel pressure regulator tune which doesn't increase longevity.  

The motor did still run great, getting me a podium at The 2019 NASA Champs and a win for my last race of the season.  However, with all of that abuse, I figured it would be a good idea to get an overall idea of the motor's health.

Engine Bay of Chelsea the Civic

Compression Test

The first test I wanted to do on the Civic was a compression test.  So, I borrowed a compression tester, and some other tools-- including a leakdown tester and an air compressor from my friend Anthony, an H2 racer, and went to town.

Ant's Car

A compression test is used to read the max pressure each cylinder can produce when the engine is cranked over.  Compression tests are useful for determining the general health of the motor.  Max pressures are recorded for each cylinder and then they are compared.  If one cylinder is down on compression relative to the other cylinders, you know that there is a problem at this cylinder.

Compression Testing a VTEC Honda Motor

If you do find a cylinder low on compression relative to its siblings, it could be caused by a variety of factors.  There could be a burned valve, bad ring, scored cylinder, cracked piston, cracked or warped cylinder head, or more.  To figure out where to begin troubleshooting a cylinder low on compression, we do a leakdown test-- this will be discussed later in this post.

Photo by Viken Photography

I decided to do the test on a cold motor with the cylinder "dry" and with the cylinder "wet."  The definitions of a dry and wet test are shown below:

A dry compression test is done with the engine in its normal operating condition.  It can be done with the engine cold or hot.  I chose to do it cold due to time constraints and because I ran into some issues when trying to do it hot (the pressure gauge broke).

A wet compression test is done to see how well the oil ring is sealing between the piston and the piston wall.  It is called a wet compression test because a small amount of oil is added to the cylinder and the compression test is then conducted.

Wet Vs Dry Compresion Test Diagram from XS650.com

Dry Compression Test

While performing the dry compression test, I made a mistake while taking compression readings on cylinder #1.  The connector for the pressure gauge was slightly loose and this likely threw off the reading.  During analysis, I will treat this reading as an outlier and assume that the average compression for cylinder 1 is actually 155.

With this being said, two things became apparent after reviewing the data:

1.) Cylinders 2 and 3 were lower on compression than cylinders 1 and 4.  Since I overheated my engine halfway through the season, it's possible that the head gasket partially failed in between cylinders, allowing for cylinder-to-cylinder leakage.

2.) Cylinder 3 was the lowest on compression of the three.  While performing further diagnostics, this cylinder will receive more attention than its sibling cylinders.

The table below shows the data from this test:

Dry Compression Test on an Acura Integra B18B1 Motor with Two Race Seasons and 170k Street Miles

Wet Compression Test

Next, I performed a wet compression test.  I added two caps of 5w-30 Pennzoil Platinum Synthetic Oil to cylinder one to start.  After attaching my pressure gauge for the first test and cranking the motor over, it became immediately apparent that two caps of oil was too much.  Wet compression at cylinder 1 came back at >300 PSI....

A "Windowed" Engine Block (Pic from Toyota-Club.net)

So, for cylinders 2, 3, and 4, I added only ONE cap of oil, using the same oil, and achieved more realistic results.  After performing the wet compression test, it was apparent that cylinder 4 had the highest wet compression and cylinder 2 had the lowest wet compression.  Perhaps the piston rings at cylinder 4 were in better condition than cylinders 2 and 3.
Wet Compression Test of a B18B1

After some consideration, I'm not sure that the wet compression test provides as much value as the dry compression test because there are too many variables in the test, such as:

1.) The exact amount of oil added to each cylinder (an oil cap is not most accurate way to add)

2.) Where the oil lands on top of the piston when it's injected into the cylinder

3.) When the oil is injected vs. when the compression test is taken (more time allows more oil to seep past rings)

4.) And more...

I think performing a dry test cold and a dry test hot would give a better subjective overview of how the engine is performing.

Leakdown test

As mentioned above, a leakdown test is used to determine WHY a cylinder may be down on compression relative to another cylinder.  A good DIY for the beginner on this process is shown in the video below:

Eric the Car Guy Makes Some Great DIY Videos

To do the leakdown test, I wanted the motor to be at running temperature to ensure the valves and valve seats were fully sealing inside of the cylinder.  I also wanted to ensure any residual oil from the wet compression test was out of the engine.  

I Ran the Civic for 15 Minutes After it Stopped Smoking

I set the air supply from my air compressor to 60 PSIG and began conducting the leakdown test, recording the leakage pressure to get a percentage leakdown for each cylinder.  I did this test twice for each cylinder.

Let's take a look at the leak down percentages below:

Leakdown Percentages for a B18B1 at Running temperature

There are four test results that show nearly 100% leakby and four tests that show realistic leakby values for a motor that doesn't run like a bag of rocks.  I treated the high readings as outliers because they didn't make sense for a motor that runs as well as mine does.  Additionally, I've seen excessive leakby on engines where the cams were still installed in the past.

With the cams still installed, slack in the timing belt, stiction between residual oil, cams, roller-rockers, pistons, and cylinders becomes a factor in ensuring the valves fully seat when the piston is rotated to TDC.  So, I took the best reading for each cylinder and used that to calculate a leakdown percentage.  

Interestingly, cylinders 3 and 4 had the worst leakdown percentages of the four cylinders, but only by 2%.  Recall that cylinders 1 and 4 had the highest compression readings of the four cylinders and cylinder 3 had the lowest compression reading.

Me and Dame Racing at Watkins Glen (Pic: Windshadow Studios)

To determine where the leakage is coming from, for each test, I listened to see if I could hear air hissing out of the throttle body (which would imply intake valve leakage), air leaking out of the exhaust (implying exhaust valve leakage), and air leaking out of the valve cover (which would suggest piston-to-cylinder leakby).

Below is a table showing where air leakage was heard for each successful test-- that is, I ignore the tests with erroneously high leakby values:

B18B1 Leakdown Test Observations

In all tests, it is apparent that leakage for each cylinder was due primarily to leakage between the valves and valve seats, which is to be expected from a motor with 170,000 street miles.  Additionally, I did hear some air hissing out of adjacent cylinders when conducting leakdown tests on the middle cylinders-- which corroborates suspicions of head gasket failure.

Oil Analysis

Oil is used to lubricate any rotating and reciprocating components in the engine, such as the crankshaft and bearings, connecting rods and bearings, piston-to-cylinder surfaces, camshafts and journals, and roller-rocker surfaces (to name a few).  Lubrication for these components is necessary because they are constantly creating friction during normal engine operation.

If adequate lubrication is not present in the engine, affected components will fail and metal bits from them will end up in the oil, in addition to other contaminants.

VTEC Valvetrains See Even More Heat and Friction Than Their Non-VTEC Counterparts

Metal isn't the only thing that shows up in engine oil when the engine is not running optimally, however.  A sample may show that an engine is running too hot as evidenced by degradation of additives or carbon chains in the oil.  A sample may also show that the engine has excessive blowby or incomplete combustion as evidenced by unburned fuel or sulfates and nitrates which have dissolved into the oil.

So, we sample the oil to get an additional snapshot of engine health from a different perspective.


Oil analysis isn't guaranteed to identify a failing engine.  For example, a healthy running engine that experiences a brief transient where a bearing is partially wiped at the crank may have a relatively healthy oil analysis, but a large, insoluble piece of bearing may be present in the oil.  This happens frequently in heavy industry with steam turbines, for example.

However, health assessments are broad and take into account many factors...we want to analyze the "big picture" and avoid relying on only one test to determine if our engine is healthy or not.

The Big Picture (A Cool and Good Stock Photo)

So, after performing my compression and leakdown test, I decided I'd send the oil to Blackstone Laboratories for analysis.  Blackstone laboratories is based out of the Midwest.  They will send you a sampling kit for free and all you have to do is pull a sample of oil and send it back to them.

A Blackstone Laboratories Oil Sample Kit

In the sample kit, they have a questionnaire that asks about the car's use, any important info like engine swaps or mods, and the age/mileage on the oil.  They have experience analyzing many types of cars from daily drivers to race cars.  They even have experience in analyzing hydraulic fluid or fluid from industrial machinery.

Within two weeks, I had my oil drained, collected, mailed, and sampled by Blackstone.  Their analysis stated that my motor was in healthy condition with wear parameters below or in-line with expected values for most motors similar to mine.  Take a look at the results below:

Blackstone-Provided Comments on Engine Health

They provide an extensive data sheet showing wear particles found in the oil with average values for comparison purposes.  The table below shows this:

Elements Found in Oil

Lastly, they provide a table on values such as percentage anti-freeze and viscosity of the oil:

Additional Oil Information

Conclusion

After discussing results from the tests above with champion level racing drivers and builders in the amateur motorsports community and after analyzing the results on my own accord, it became apparent that the motor was in good health.

Wear material in the oil was low, suggesting the engine did not have significant degradation due to a lack of lubrication anywhere in the rotating/reciprocating assembly.  Leakdown wasn't exceptionally high, and any leakdown present in the engine was most likely due to valve-to-seat leakage, and not bottom end leakage which would require an extensive rebuild.

The head gasket was a likely contributor for high leakdown % on cylinder 3 and also for lower compression numbers on cylinders 2 and 3.  This was likely the result of damage from pegging my coolant gauge high during a qualifying session at the beginning of the season in 2019 at NJMP Thunderbolt.

Me and Ken Racing at Thunderbolt (Pic: Windshadow Studios)

Overall my B18B1 is in good condition.  I know that I can likely run this motor for another season with no issues.  This knowledge will allow me to focus my efforts on other upgrades to improve the car's speed, such as suspension and brakes.

Thanks all for reading.  I hope this post was informative and interesting.