Bill Hoddinott: Tom, Betty and Gene have given us the outlines of the streamliner, now would you give us the more detailed design concepts. I think the car is of intense technical interest!
Tom Burkland: Thank you for saying so, Bill. The basic approach with using two 450 inch Donovans was to make sure we had plenty of power for very high speeds. You see, of course, that we could have two engines in the same frontal aspect as one and that part of the design is SO important.
An inherent benefit is that two very powerful engines are going to be operating in a low-rpm, low-heat, low-stress mode here which obviously helps their endurance and longevity. Thus you don't have to work on them after every pass out at the Flats. And we have had them virtually trouble-free over the years.
Bill: Funny you should mention that, Tom, because this harks clear back to the Schneider Cup seaplane races of the 1930s when the V-12 engine was so ideal because it had such a slim small frontal aspect, not much bigger than the pilot. The prop spinner and the front part of the fuselage just covered the engine.
Tom: Exactly. But having said that, it later turned out that, due to wheelspin, we have never been able to use more than about 60% of our power. So we think now that we might have been able to get just as much speed with one blown methanol Donovan working closer to 100% power. This would have shortened and simplified the car a lot and reduced expense and weight. The inherent cooling characteristic of methanol in correct mixtures would still probably have retained the endurance the engine must have for the long pulls at Bonneville.
We chose the Liberty five-speed Pro Stock gearboxes because they had a rated torque capacity of 940 foot-pounds and had been very successful with heavy drag cars on pavement. We thought the lack of traction at Bonneville meant the boxes would never see any torque inputs near that, since the wheels would slip first.
We knew we needed a lot of gear ratio to cover such an enormous car speed range and first gear is 2.6 in the Liberty, and the rest of the ratios are close, with 1 to 1 in top gear. Another attraction of the Liberty was that plenty of spare parts for it were on the shelf at the manufacturer..
However, we did experience some breakage during acceleration with this gearbox at the beginning, and we finally realized that it was being caused by the heavy shocks and pounding you get when the car is running over a washboard surface and the wheels are spinning to an extent - when there is no form of shock absorber in the driveline, like the typical spring assembly in a common clutch plate. You realize the car needs to reach 300 mph from rest within nine to ten seconds to make a normal run at Bonneville.
Liberty did supply us some stronger wider gears later made from better materials on larger diameter cluster shafts, and we got some custom high strength shafts made. And, we adjusted our McLeod Long-style clutches carefully to get a calibrated slip during gear changes with the air-shifter. All these measures resolved the gearbox problem to manageable levels.
Had I to do it over again, I would have used B&J three-speeds, since three gears would have been enough. The engines have so much torque at all rpm ranges, like all blown Hemis, that gears two and four in the Liberty are really superfluous. The softer shifts of the planetary style transmissions could even improve the driveability on the track.
Bill: What are some of the considerations about the blowers and mechanical fuel injection?
Tom: Before he died, Gene Mooneyham made us two sets of 8-71 blowers with magnesium cases and rotors. We'll always be grateful to Gene for that because he had to go to a LOT of trouble to get parts for, especially, the two for the front engine that are reverse-rotation. We wanted magnesium blowers, injector throttle bodies, blower manifolds and valve covers for the Donovans to reduce the top weight. Wanted to keep the CG of the car as low as possible but even with all the magnesium parts, CG is still 17 inches off the ground which gives a bit of uncertain handling with the "tricycle" wheel configuration. When people ask me how the car handles, I always say, "Like a loaded cement truck!" Not THAT bad, really, but unlike the Datsun, which I could put between orange cones on the Flats eight feet apart at 300 mph, I think I would have to set my cones at 40 feet to guarantee getting the streamliner between them at maximum speed!
But we came to regret all the magnesium parts up top eventually, and I wish I had taken the extra 50 pounds of aluminum parts.
Tom: Because no matter what kind of coating you put on the magnesium castings, the salt at Bonneville is ALWAYS trying to eat them! It's much worse than aluminum and a never-ending battle.
We found out later that larger blowers were not required to make adequate power, either. We have never seen more than 11 psi in our blower manifolds since as I say, we have never been able to use all the power we have due to the wheelspin factor.
Many years ago, Crower made an eight-hole throttle body with a compact integral barrel valve and eight-outlet distribution block for the 8-71 blower and we acquired a few of them. This is ideal for our application because it gives a slow-opening characteristic for your throttle. We were after DRIVEABILITY for our streamliner and this was one way we could get it.... These Crowers are also very low in vertical height which was important for our overall engine dimensions that as I said, was the first design consideration for the car body.
Now Bill, you understand that for drag-racing and most Bonneville cars with mechanical fuel injection, all people need or want is two positions of the throttle. Closed for startup and idle, and wide-open.
For this need, a simple form of injector is often sufficient, but people sometimes use a high-speed bypass to overcome the inherent richness you get with higher rpm because fuel flow increases on a straight line graph, and air flow into the cylinders drops off as rpm increases, due to reducing volumetric efficiency.
In our case, we knew we would have an overpower situation especially during the acceleration phase, and as it turned out we could only use about 5% of throttle in first gear or wheelspin would be excessive. We could feed in more throttle progressively going up through the gears, but even in top gear never more than about 60%. And you need delicacy of throttle control to modulate wheelspin and fully utilize the available traction.
So knowing we would be dealing with this, I spent a whole week of 12-15 hour days at Tak's shop that Gene mentioned before, at the flow bench. First thing was to get a matched pair of Enderle fuel pumps, so that the two engines would be on the same page as far as fuel flow. Next I designed my own spool for the barrel valves, with a very slow-opening flat on it to meter the fuel on the tip-in very carefully. Then we added a metered secondary bypass ramp for part throttle fuel curve control. The small plates in the eight-hole throttle body help the exact air metering at tip-in rather than almost full air-flow happening as soon as you start to step on the pedal with a typical large throttle area injector.
Next thing is we decided to put about 30% of the fuel into the throttle body nozzles, and the rest into port nozzles in the blower manifold. Our blower manifold is a low-profile type to keep vertical engine height low, but this means a sort of flattish plenum under the blower so that air flow to the head ports is bound to be somewhat unequal to the various cylinders despite the manifold pressure. Add to this the effect of acceleration on the fuel spray and vapor coming from the top hat nozzles, which could be biased to the rear of the manifold, AND the fact that the bulky 8-71 blower does not sit on the fore-and-aft centerline of the engine. The front engine is also sitting backwards in the car so acceleration impacts on the liquid fuel distribution are reversed in comparison to the conventional rear engine installation. So there's a lot going on here and to ensure that each cylinder gets the most perfect fuel-air ratio at all rpms, the size of the port nozzles is varied to achieve balance between the 16 cylinders. Even though not all cylinders will get the same amount of air, what air they do get will be mixed with the optimum amount of fuel for best combustion and hence power.
We have an exhaust gas temp thermocouple on each cylinder's header pipe as a check on combustion in it, which is recorded by our Cygnus data acquisition equipment. I should mention that my computer wizard grandfather Keith Hunter purchased the original Cygnus equipment. This system, which was rather expensive for its day, greatly reduced the development phase of the project and served as a valuable driver training device. The EGT validates the size of the nozzles you have at the ports.
The engines idle at 1800 rpm and in this mode they run on the top hat nozzles only.
Bill: Since methanol by itself does not vaporize very well for starting under 50 degrees F, how do you get the engines fired up on a cold morning, say at World Finals?
Tom: We have some little rain-gutter-like troughs in the air intakes for each engine and we squirt a calibrated amount of gas into them from a squeeze bottle which runs down to the throttle bodies, into the blowers and down into the blower manifolds for a prime for starting. This works fine, but you do have to be VERY careful with this procedure because with raw gas exposed, if you had a backfire when cranking or on startup which managed to get up past the blower, you could end up with a fire in your air scoop.
You have all this for a cold startup, but during the day at Bonneville your engines will start on the methanol okay just primed by holding the throttle open a little which enables more fuel to enter the blower.
Bill: WOW, Tom, that's about the MOST sophisticated mechanical fuel injection system I ever heard of!
Tom: Thanks, Bill. It works, too. Probably the best statement on the fuel system was when we were accused of running an automatic traction control in the first few years of the development program before they were legalized in the rule book.
Bill: How does the air scoop on top of the car figure into the fuel injection and blower arrangement?
Tom: It is designed for minimum size and drag but to give ram air to the tops of the throttle bodies, which are sealed to it. Despite the volume of air the bodies take in there is 3.5 psi over them from the ram air at 400 mph car speed. You see the scoop has a divider in it to ensure that each blower gets an equal share of the air and it is equally distributed over its length through the eight throttle bores on each engine.
Bill: I think that covers the intake side, what are the exhaust system specs?
Tom: 2-1/2" primary pipes into a 5" titanium tailpipe for each side of each engine. The exhaust systems are ceramic coated both inside and out by HPC, then wrapped with stainless steel insulation blankets, and they are in bays separated from the engine bay; all to minimize the heat factor in the engine area we mentioned before which initially caused trouble on coastdown. The cooling air from dedicated external inlets is distributed into these header bays on each side of the car through internal ducting and exhausted through a venturi extractor around each tail pipe at the rear. The engine bay cooling system also includes a set of spring-loaded exhaust doors on the center bay to vent heated air after the car comes to a stop.
Bill: I think most people will be aware of the engine construction, as noted the Donovan follows the 392 Chrysler Hemi except for a lot of improvements in strength, the dry decks and the wet cylinder sleeves. It is worth mentioning that the massive one-piece aluminum girdle on the Donovan which forms the main bearing caps goes between the block proper and the oil pan and is visible in some of the photos you provided.
What compression ratio and ignition lead do you use for the blown methanol engine?
Tom: 10.5 compression and we started with 36 degrees fixed timing, but later dropped it back to 28. We had so much power it was possible to retard the timing a little, which reduces the combustion pressures inside the engine and also makes it easier to crank it on the starters. The more spark lead and compression you have the more the pistons fight the starter. If I had it to do over again, I would have gone with a lower compression ratio as well, to ease the stress on the starters. That would also reduce the peak pressures inside the engine and probably help the driveability issues.
Bill: The cooling system you have is very interesting, and you have what you call a "surge tank" at the top between the engines. How does it all work?
Tom: The surge tank at the top is the filling point for the whole system, and it is atmospheric vented. The nose cone is our water tank, holding 58 gallons, and it has its own manual purge valve at the top to allow trapped air to vent during and after filling. The nose cone remains essentially 100% full during all operation. There are two T-33 jet aircraft fuel pumps in the lower portion of the rear bulkhead of the nose tank, one for each engine. They are 24 volt units used here for water pumps and powered by two deep cycle 12 volt golf cart batteries on board, which also supply all the gauges and other onboard electrical requirements. The data recorder has its own dedicated battery system to avoid any electrical noise issues.
The water pumps deliver to the engine by one-inch hoses and aluminum hard lines under the cockpit and the outlets from the engines go to the surge tank. The outlets from the surge tank are below the working water level inside so the suction in the nose cone from the pumps continuously pulls water back to the nose cone tank.
Bill: Okay, the cooling system is vented at the surge tank, so there is no pressure system here like an ordinary road car.
Tom: That's right, and there are a couple reasons a pressure system is unnecessary or would not work. First, the nosecone metal shell is not thick enough to withstand pressure, and would distort. This tank is actually the exterior surface of the car with finish paint applied directly to it. Next, although we have maybe eight psi in the water jackets of the engine due to the restriction in the outlet plumbing and surge tank, we don't want any more because the big o-rings around the bottoms of the thick removable iron cylinders of the engines tend to seep a little as it is. Water in the oil is not too good for bearing life in these highly stressed engines! Finally, we don't see high coolant temperatures at Bonneville with the current tune up and power loading.
Bill: A very clever system, and what temperature do you see in the tank at Bonneville?
Tom: About 160 degrees F is the max we have seen. Due to the large volume of water and the assistance of the methanol fuel cooling effect. You know on a cool morning at Bonneville, even in the low humidity there you will see frost form on the throttle bodies and blowers when you make your first startup of the day. If it was in a high-humidity environment like you have in Virginia, there'd be a quarter inch of ice forming over the entire exposed intake tract! The water temperatures are displayed in the cockpit by Autometer mechanical capillary tube gauges with a high temperature warning light system that comes on at 220 degrees.
Bill: I think the dry sump oiling system you have for the engines is also very interesting. It's not the timing-belt-driven type usually seen. Why a dry sump in the first place?
Tom: There are several reasons for the use of a dry sump system here, Bill. To start with, it's safer because the engine only has about two quarts of oil inside it when it is running, so if you have an engine failure where oil might get on the headers and catch fire, there is less fire to contend with. It's also a lot less slippery stuff under the rear tires to slide around on, and clean up for the rest of the racers.
Next, we wanted to minimize the vertical height of the engines, which of course includes the oil pans, for frontal aspect and CG considerations. The oil pans are custom built to match the smallest diameter clutch can and clutch assembly we could make work for the amount of torque being produced.
Another thing is we can use large oil tanks which keep oil temperature down. Here we have four gallons of oil per engine and tank, we use 60 Wt Pennzoil Racing oil which is recommended for blown fuel engines, and its temp never goes over 210.
Bill: What about the pumps?
Tom: We have a Barnes two-stage pump in the normal place inside the oil pan. One stage supplies the pressure side of the engine, the other scavenges part of the pan and sends oil back to the tank. But
that by itself is not enough for complete scavenging. You need about three times the scavenge capacity as delivery, since the return oil is foamy and you are pumping a lot of air with it. So we have a two-stage Aviad scavenge oil pump for each engine, driven by the camshaft, which sucks the oil from certain points in its oil pan and sends it back to the tank. The fuel injection pumps mount on the front of these Aviads to receive their drive.
Bill: Is there ever any issue about the Aviads, being up high above the oil pans, being able to prime themselves from a cold startup and start functioning?
Tom: No, they are designed to operate this way and the whole system works fine. Gene fabricated the oil tanks from aluminum to fit available spaces forward of the engines, with big hoses and fittings so gravity assists the delivery of the oil to the Barnes pumps during acceleration.
Bill: Are the aluminum oil tanks rubber mounted? I'm thinking here of the violent vibration to which the car is subjected.
Tom: No, they are mounted by tabs metal to metal but well supported by the chassis tubes upon which they sit. We've never had any trouble with leaks or cracks with any of the many custom tanks in the car. More of that Gene Burkland welding wizardry!
Bill: What kind of oil pressure do you see with this setup?
Tom: On a cold startup, 110-120 psi depending on the temperature, and at the end of the Long Course, with everything hot, maybe 80 psi. This is more than enough to ensure bearing life in the engines. A low oil pressure warning light system comes on at 50 psi to alert the driver.
Bill: Fuel engines normally contaminate their oil quite a bit and turn it milky, and people change it all the time. Do you have that?
Tom: No. That characteristic, I think, usually arises because people use very rich mixtures of methanol and/or nitromethane, which results in washdown past the piston rings. In our case, we don't see that because our fuel/air ratios are so tightly controlled from the ideal level through all rpm ranges. So we don't have to change the oil after each run. The FIA one hour turnaround would be virtually impossible to achieve if an oil change was required between the two runs. The reservoir tanks have dip sticks that allow us to monitor the oil condition and level to make the changes when necessary.
Bill: Okay, I think we have covered the powertrain aspects of the car pretty well between this and what we have in previous Parts but would you describe in detail the front and rear axle spiral bevel gearboxes.
Tom: Sure, we originally thought we might have some trouble with these components since they are highly stressed, but on the contrary, they have been completely trouble-free. The spiral bevel gearsets for them are 1 to 1 ratio and are aircraft quality parts originally manufactured for helicopter tail rotor gears. They run half-submerged in two quarts of synthetic gear oil inside their iron boxes. In manufacture the two gears are lapped together to provide an ideal fit and they are then indexed for correct assembly. With the 1 to 1 drive the same teeth are always in contact and they take up an ideal finish with maybe .004" backlash for oil clearance. The torque capacity of these final drive gearboxes is over 2000 foot-pounds. After the 2001 crash we stripped, inspected and magnafluxed everything about these gearboxes and found no defects whatever, so reassembled and continued to use them.
One point I might mention is that at the design stage, I would have liked to have the two rear wheels even closer together than the 13-inch tread centers we have here. It would have allowed the body to be narrower in that area leading into the tail flap. I considered trying to put the drive into one side, but it was just not practical to configure any other way but with the driveshaft going into the gearbox between the tires, so that's what we have.
Bill: For the next part, let's cover what the driver's experience is like operating the car.
Copyright © 2009 Bill Hoddinott