Update: IAT2 test results added

Today I performed an intercooler pump test on the Jabsco 50840 12v pump as a replacement for the Bosch OEM intercooler pump in my Cadillac STS-V.  The Cadillac STS-V uses the Bosch 00 392 022 006 pump.

My Cadillac has a 2nd heat exchanger, an S3 provided by Timmy C, and uses a 1 gallon AVS inline intercooler tank.  The goal of these tests is to improve intercooler cooling, which will enable the supercharged V8 to make more power.

Test setup with Jabsco pump in the loop

Intercooler pump test setup

To run a bucket test I put a FROM bucket at intercooler height, flowing into the pump and system, and out from the intercooler to a TO bucket.  This is similar to my previous intercooler pump test setup, for continuity.  I have added a TO bucket that is a mixing bucket which makes it very easy to measure output.

Predicted intercooler pump test flow:

Pump rated flow vs head pressure and system pressure curve

The Jabsco centrifugal pump does 29 gpm against no resistance, and the Bosch pump does 8 gpm against no resistance. In an actual system the pump curves overlaid with the system resistance curve predict where the pump will actually run in that system.

Intercooler Pump Test Results

In today’s test the Jabsco pump measured at a steady 2.25 gallons per 30 seconds, or 4.5 gpm (gallons per minute).  Previously, the Bosch OEM pump ran at 3.5 gpm, which would suggest the system pressure at that flow was 6 psid.  As flow increases, the resistance pressure in the system increases roughly at the square of the flow.  I expected the Jabsco pump to run 4.3 gpm, so the 4.5 gpm is good.  The Jabsco has 1/3.5 = 28.5% more flow than the OEM pump.

I still have more work to do to complete pump installation.  I will retest with the pump in place.  I will follow-up the intercooler pump test with IAT2 tests to see how the new higher flow impacts intercooler cooling.

Continuing to modify and fit check the Jabsco pump on the OEM bracket.

Intercooler Pump Test Update

Today I completed the pump install and conducted a test run with the new pump:

Intercooler pump test showing IAT2 over time during a 20 min drive

Click on the graph for a larger version.

What this graph shows is temperature in degrees F on the Y axis, over a 20 min drive on the X axis.  Spikes in the graph represent acceleration runs.  The red line is the OEM pump without the inline tank; the blue line is the OEM pump with the inline tank; the green line is the Jabsco pump with the inline tank.  Today it is 92F ambient; on the prior test days it was 90F ambient air temperature.

I read this test result to show that the jabsco pump’s greater flow allows the system to reach an IAT2 equilibrium temperature of 118-120F, an improvement of 7F.   The acceleration spike is a 20F increase, an improvement of 5F in addition to the lower equilibrium, so that at peak temps are 12f lower than previous hot day.

Intercooler pump test conclusion:

The Jabsco pump has a 1 gpm higher flow than the OEM Bosch pump.  This 1 gpm greater flow lowers the operating IAT2 temps by 7F

Ambient temp 92F; ambient air temp resets to 32f until the ecm can determine actual temps. Near gear change at top of 2nd gear.  Losing 1 degree of advance to IAT2 retard.

Today I conducted a new flow test on the intercooler cooling system for my Cadillac STS-V.  Previously I ran a bucket test to determine the actual flow through the intercooler cooling system.  Since that test, the system has been modified to add an inline intercooler tank.  I plan to change the pump from the OEM Bosch pump to a Jabsco 80540-12 pump soon, so today’s test confirms the flow rate prior to the next mod.

The overall purpose of this series of tests and changes is to improve intercooler cooling, which will result in lower intake air temperatures after the supercharger, and let the engine make more power.

To run the test I place a ‘from’ bucket on a table, which has replaced my earlier box due to box degradation.  The From bucket flows into the first heat exchanger.  The To bucket is flowing out of the intercooler.  The goal of the test method is to measure the flow as if the test were not going on — to not impact the flow.  This is why the From bucket is in a raised place.

Today my Son assisted me with timing and checking the test setup.  We ran the test with the car engine on, to simulate working conditions for voltage.

Intercooler cooling flow Result:

We measured a flow of  1 & 3/4 gallon per 30 seconds, or 3.5 gallons per minute (GPM).  This is identical to the flow measured in the earlier test without the inline tank.  I conclude that the inline tank presents no more resistance to flow, or that the changes in hose routing during the tank install improved flow sufficiently to offset the resistance of the tank.