This article originally published in Issue 61 of the Turbo Diesel Register.

Technical Topics

IT'S ABOUT THE (fuel) ECONOMY STUPID! (page 4)

VENDOR RESPONSE TO THE NEED FOR TEST DATA

Your closing comments:

TST: TST has been in the business of increasing power and torque for a decade. Up until 2006 about 90% of incoming started with, “How can I get more of that power and torque stuff?” Then, almost overnight, the question became, “How can I get more mileage out of this big beast?”

Power and torque increases were always easy for us to measure as we test on our own chassis dynamometers. Typically, we would leave one of our test trucks on the dyno for 6-8 weeks at a time, daily trying a program change or parts change, and let the dyno tell us if the engine liked or disliked the change.

My first job in Cummins engineering in 1966 was keeping track of hundreds of test semi-trucks running without ever changing motor oil. Monthly, I would have to pull oil samples on each of these trucks, record mileage, any oil addition since the last check, and document the results of the oil analysis tests. I started keeping record books on my personal vehicles at that time, recording every event, fuel fill, oil change, new tires, etc. As a part of the personal record keeping, I’d calculate the mileage at each fill and noticed how the mileage constantly changed tank to tank. I’ve continued this practice to present day with my diesel trucks, keeping Excel spreadsheets to show each tank’s mileage, running average, and change in fuel cost. Carefully filling the tank to the top, with the aid of a tank vent kit, still did not eliminate the variation tank to tank. My Excel spreadsheets (and your notepad and pencil) are good for long-term trends, but I view them as consistently inconsistent for short-term evaluations.

With customer requests for better mileage becoming the number one priority, I spent many hours wondering how TST could evaluate fuel economy without burning several tanks of expensive fuel. I recalled my experience in the Cummins Engine research labs where we often monitored the fuel consumption of running engines without the hindrance of an attached vehicle. All engine manufacturers gather data to calculate brake specific fuel consumption (BSFC) in order to compare the relative economy or efficiency of various engines. Typically an engine would be run for several minutes at a constant brake horsepower and the fuel used was measured in pounds. The word brake in this case meant the engine dynamometer which measured flywheel torque and engine rpm such that engine flywheel horsepower could be calculated. A simple calculation could be made with the gathered data by dividing the weight of the fuel used in pounds per hour by the horsepower being generated. BSFC number like 0.350 pounds per horsepower per hour was the end result. A very good engine might run a 0.330 BSFC while a poorer engine might run close to 0.410 BSFC.

I felt finding a BSFC-like number for a Turbo Diesel would be a good way to check the relative fuel consumption. I knew our chassis dyno could be set to measure the horsepower at the wheels. But at what horsepower did we need to run the tests? The TST staff brainstormed the question and decided we needed to determine the actual horsepower it took to move a Turbo Diesel pickup at various road speeds. We took our ’03 Dodge Ram out on the Interstate highway and ran it for several miles in both directions while recording instrument readings for air temperature, mph, gear selected, turbo boost, egt, rpm, and rail pressure. We gathered data for road speeds from 55 mph to 75 mph, in 5 mph increments. We then tied that truck to the chassis dynamometer and tried various loads until we could reproduce the instrument readings we took out on the Interstate. This gave us a horsepower value to use in our fuel economy tests for various road speeds. (See figure 1.)

Next, we needed a way to accurately measure the fuel used during our testing. While we considered trying to carefully refill the stock Dodge fuel tank after each run, we quickly discarded that idea because it was impossible to fill that tank to the same level each time. We also considered placing a small tank in the bed of the truck that would be easier to refill to the same level each time, but that still left us with the problem of measuring just how much fuel it took to refill the tank. Suddenly it became obvious, let’s run the truck using a remote fuel tank that sets on a very accurate scale and simply weigh the fuel as it is consumed. By carefully weighing a gallon of fuel, we could then determine the number of pounds per gallon. We could then use this pounds-per-gallon number to convert our pounds of fuel used back to gallons for the mpg calculation. This is the procedure that is used by Cummins in official ASE-type testing of big rigs at fleet accounts. We purchased an electronic scale with a digital readout that was guaranteed accurate to one hundredth of a pound and used a transparent five-gallon plastic jug as our fuel tank. We unhooked the quick connects from the truck tank and plumbed them such that the engine would draw and send its return fuel to the plastic jug (See picture below).

Ready to start testing? We thought so. We brought the truck up to 55mph on the dyno and set the truck cruise control to hold the speed. That part worked great. Next we started increasing the load on the dyno until we could repeat the horsepower number we found from the earlier Interstate highway testing. That part worked great too. Then we drove the truck for five minutes at these conditions taking the fuel weight before and after the test. That part worked great too. Well, almost. One of our technicians accidentally touched the five-gallon jug and realized it had become very hot from the engine’s returned fuel.

It took about four hours to add 50 feet of copper tubing to our return fuel line. We dropped that copper line into a large cooler and returned the fuel to the plastic jug. We then used a garden hose to run cold city water through the cooler to keep the returned fuel cool. By regulating the city water flow we could maintain the fuel temperature at a usable level.

We were ready to start testing again. This time we were able to repeat the weight readings time after time during the five-minute test. We now felt we had a way to quickly and accurately determine how various engine changes would affect fuel consumption.

Back in 2002-2004 we developed the PowerMax CR for the 5.9 HPCR engine. We engineered a way to vary injection timing, injection duration, and rail pressure on the fly. Of course our motive was different back then; we were after the most power and torque at the lowest exhaust temperature. Now with a new goal of best fuel economy, we started all over again trying to optimize timing, duration, and rail pressure.

First we determined the fuel consumption curve for the stock engine using our new measurement method. (See lower curve in figure 3.) The mileage numbers from this test were quite a bit better than our record book showed for this truck, but keep in mind that our new test method was steady state only, no starting and stopping. I figured if we could improve the steady state numbers, mileage would also improve on the highway.

To start, we tried the injection timing schedule from the PowerMaxCR that gave us the best power curve. That timing helped fuel economy a bit, but not a significant difference. Over the next few hours we tried many different injection timing settings and selected the ones that gave us best economy from 55 to 75 mph. (See upper curve in figure 3.) We then tried varying the rail pressure while using only the best timing found earlier. Changing the rail pressure didn’t help.

We added an “Economy” setting to the PowerMaxCR as a result of these tests. To date, customer experience has been mixed. Some claim big gains like 3 to 4mpg, some report no change at all, while a few claim they lost mileage. How could this be? Looking back, all testing was done on the same truck. The truck was a ‘03 Quad Cab, 4x4, SRW, long bed, six-speed manual, 3.73:1 axle, stock BFG tires (LT 265/70 R17). The only modifications to the truck were cab high full length cap, FASS HPFP 95 gph pump, and a South Bend Double Disc clutch. I felt the FASS pump and SBC clutch had no affect on the mileage test, but were needed on the truck for full power testing done separately.

What next? Let’s run the rail pressure box but with stock timing. Bigger injectors get the fuel in quicker, so do they help the same as advancing timing? How about a 48RE automatic truck? Is there any difference in the behavior of the ’04.5 HPCR engine with its 325hp? I’ve got a 6.7-liter chassis cab to test. Then there is the 6.7-liter pickup with the terrific new 68RFE six-speed automatic. What? You want me to run an ’89, too. It doesn’t end, does it. Maybe next issue!

Mark Chapple
TST Products