Holy Cow Tractor Pulling Team

How do you connect the engines front to rear?

A little duct tape and a little super glue. Seriously now, this is hard to describe in words, so I took some pictures during my last rebuild.

I am going to start with the rear engine first and work forward. That is how they get bolted on the tractor so it makes for a logical progression. First we start with the crankshaft. The first thing is to machine 4 keyways in the snout of the crank. Stock Chevy cranks come with with a single 1/4" keyway to position the harmonic balancer. (Well, there is also a key for the timing gear, but let's ignore that) Some, if not most, high performance cranks come with 2 keyways, spaced 180 degrees apart. On a stock crank, the belt pulleys attach to the snout and drive the accessories like the alternator, water pump, etc. On a high performance engine, all that crap goes away and the only thing driven off the crank is the blower. My blowers take about 400 hp to turn, so there is a lot of load on the crank hub, and that is why 2 keyways are installed. Well, in a crank-to-crank application, you are now transmitting the power from another engine through that same hub, so the keyways are doubled again. I built a fixture and cut extra 2 or 3 keyways myself, but any crank shop could do it easily. Once the crank is machined, the bottom of the engine is assembled as normal.

Next, the blower hub is pressed on. I place one key in the crank, apply a little Loctite shaft retainer, and draw the hub on, first with a hub installer and finally with the hub bolt. Once the hub is on, I drive in the remaining 3 keys with a hammer and a punch. Most of my hubs came from Ralph Banter, but Fowler Engines in Columbus, OH also sells them. Regardless of whether I am building a front or a rear engine, I always build them the same. You never know when engines will be shuffled around and what was a front is now a rear or vice versa. The washer on the front of the hub is also custom made. I has a protrusion on the back side that almost touches the tip of the crank. This keeps the keys from walking forward. As a result of the Loctite and the 4 keys, the hub is difficult to remove. I made a puller that has a central post a little smaller than the diameter of the snout. The post is welded to a plate that has six bolt holes the same spacing as the blower pulley. Six bolts are threaded in and tightened a little at a time until the hub pops off.

Now the engine assembled the rest of the way. The next step is to attach the blower pulley. The pulley is slid onto the hub followed by the rear coupler flange and bolted on.


Once the hub is installed, the shaft can be inserted. The next thing to notice is that splines have been cut off the center of the driveshaft. The length of spline removal is just over the length of the coupler flange. This allows the flange to slide back and spin freely, which is important when synchronizing the engines. Lastly, the blower belt is put into place. Quickly, the belt will become captured and if this simple step is neglected now, it will require a lot of work later on.

So far so good. Now it is time to finish the job. The left and right brackets are bolted to the engine block. The brackets are used to bolt the two engine blocks together. The important thing here is that the centerlines of both crankshafts are absolutely in line with each other. If the centerlines are not concentric or not parallel, there will be rapid wear at the coupler. The drive shaft is coated with anti-seize to act as a lubricant. This reduces friction and improves the life of the components, but has a downside. When the engine is started, excess anti-seize is thrown off onto the inside of the pulley. You can't see it, but it is nasty if the coupler has to be removed. The front coupler flange is slid on. Finally, the upper shield is bolted into to place. NTPA requires all driveline be shielded. We use a two piece shield to that coupler access is possible. Basically, the upper shield is never removed. So that is it, the rear engine is now complete.

The next step is to install the rear engine in the chassis. The clutch is installed, which supports the rear of the engine and the front is supported underneath with a jack. Now the front engine is installed. Studs on the bell housing mount are aligned with the holes in the brackets and bolted down. The jack is removed and the front engine is bolted to the frame. Remember, the frame will bend and twist, but the engines can't. So, the rear engine is mounted solidly to the clutch surface. This both supports the engine and prevents the reaction torque from spinning the engine. The reaction torque from the front engine is transmitted though the brackets to the rear engine. The front engine mount only supports the weight of the engine. It allows the front engine to move side to side, front and rear, and to twist relative to the chassis. Once the engines are bolted on, the cranks are positioned. While I enjoy sharing the technical aspects of this sport, I am not going to divulge competitive information. The synchronization of the engines I think it one of those areas. I will say this, some guys just slap them together, some place the cranks in the same position (Front #1 TDC to Rear #1 TDC), some place them 360 degrees out of phase (Front #1 TDC to Rear #6 TDC), some 360 +/- 45 degrees out of phase (Front #1 TDC to Rear #6 45 BTDC or 45 ATDC), and other methods. However I do it, I place the front and rear cranks in their respective positions (guess that means I don't just slap 'em together).

Now you crawl under the engine and slid the front coupler back onto the unsplined section of driveshaft. So here is a little bit of magic. You want a number of splines that is not evenly divisible by the number of mounting holes. So, I have 16 splines on the shaft, which means 22.5 degrees per spline, or I could align the engines +/- 12.25 degrees. But, there are six bolt holes, or 60 degrees per hole. However, if I move three splines (67.5 degrees), but back up one hole, now I have moved just 7.5 degrees, or +/- 3.75 degrees of alignment. Since 16 and 6 have a factor of 2 in common, we could have done better with a 17 spline shaft. Since 17 is a prime number, it has no factors in common with anything and could have achieved +/- 1.7 degrees of alignment. However, if 18 splines were used, which is evenly divisible by 6, then only +/- 10 degrees of alignment could be achieved. And you thought all that Least-Common-Multiple and Lowest-Common-Denominator stuff you learned in 4th grade was useless! OK, back to the real world. You spin the coupler and get as close as you can and then bump one engine one way or the other for a perfect fit. So, I may be off a few degrees, but under full power, I'll bet the crank twists more than that anyway.

Now, the front flange is bolted to the clutch mounting surface of the front engine and the shielding is bolted in place.

Now, remember that blower belt I mentioned earlier. What if you have to change it? Getting it off is no problem, since usually the only reason you would replace one is because it broke. But, both the drive shaft and the fuel pump go though the center of the belt. First, remove the fuel pump, then remove the lower shield. Both the front and rear coupler are unbolted and slid together. This gives enough room to drop the shaft and both couplers out the bottom in the gap between the blower pulley and the front crank. A new belt is slid on, and then reassembled in the reverse fashion. It takes about 5 minutes to change a front belt and about an hour to change a rear. Some guys wad up a spare belt in there, so if the first belt breaks, the spare is ready to be installed. However, that requires back bending the belt, which is never a good idea. I have also seen the spare belt come loose and get into the blower pulley and damage the belt, before it is even run. The other thing I have seen is the engines mount in a cradle instead of using brackets. Then, the front engine can be slid forward to drop out the shaft. The cradles have some advantages, but also weigh a little more.

How do you connect the engines side to side?

It is not magic. There is a cross box. The box has a drive shaft coming in from the front on the left and right sides and a center shaft coming out the rear. Each pair of engines is mounted with a clutch that connects to one of the front drive shafts. The clutches are slipper or centrifugal clutches. As the RPM goes up, the clutches gradually engage until they are fully locked.It is the same principal as a chainsaw or a snowmobile, but a different design (and bigger). Once the engines reach 3000 RPM or so, the clutches are fully locked and the there is essentially a solid connection from the crankshaft to the drive shaft, into the cross box where torque is combined with the other side and out the center shaft, through a transmission, and into the pinion in the rear end. Without a doubt, this is the most trouble free system of the entire tractor.

Is that a chain in there?

No, five gears. See the picture below.

There are 5 gears, indicated in the picture by the solid circles. You are looking at the front side of the cross box. The far left and far right gears (above the Gottman stickers) attach to a stub shaft that gets coupled to the left and right clutches through a driveshaft. Those gears both spin clockwise from the perspective of the picture. The next two gears inside (the two highest gears) are simply idlers. They spin counter-clockwise. The center gear spins clockwise and has a stub shaft sticking out the rear that is coupled to the transmission. The four outside gears are all the same diameter and tooth count, while the center gear can be a different diameter. If it is smaller, then the gear box will be an overdrive box (output spins faster than the input). If it is larger, then the box will be an underdrive, and if it is the same, then input and output shafts spin at the same speed. Each different ratio requires a different cross box because the center to center spacing of the gears changes. I have 2 gear boxes, a slower ratio when I run the state and regional national circuits with 3 engines and a faster ratio when I run grand national with 4 engines. The only difference between the two boxes is the tooth count on the center gear and the position of the center of the idlers. The distance from input to input is the same and the rise from input to output is the same, so I can just unbolt on and bolt in the other. Additionally, if I were to run 5 engines, the center gear would be smaller still with two additional gears stacked of top of it going up to a third input shaft for the 5th motor.

How do you start it?

I can't without some help. Because I pull in a 7,500 lb class, I don't have enough "extra" weight to carry a starter and batteries. One of my crew attaches a quick connect starter to the front left blower pulley. A neutral safety switch is connected so the tractor may not be started in gear. Lastly, the starter engages to the blower through a wedge shaped 3 jaw chuck. This design allows the starter to apply tremendous force on the blower, but when the engines starts and the blower is turning faster than the starter, it gently pushes the starter jaws away to avoid over speeding the starter. To prevent damage, the crew person must ensure the jaws are engaged before applying power to the starter. The starter is connected to 3 commercial batteries in series, giving 36 Volts of power to the starter. Both the starter and batteries are carried on an ATV or Wagon.

On my signal, the crew person squirts some methanol in the top of the blower. I make sure the throttle plates are wide open, to ensure the methanol goes into the blower, and does not run out. When the engine is shut off, the fuel lines are purged. This means there is no methanol available to start the engine. Instead of cranking on the engines for a long line to refill the fuel lines, a small squirt of methanol gives the engine something to burn until it can become self supporting. I then close the throttle and make sure all 4 fuel shut-off valves are open.

Again, on my signal, the crew person flips the switch which engages the starter. The magnetos are grounded to prevent the spark plugs from firing. I allow the engines to spin a bit to build momentum so they don't kick back. Since magnetos doesn't have any advance mechanism, they are set with a fair amount of advance, the spark plugs fire well before the piston is at top-dead-center. At the slow cranking speeds of the starter, the engine wants to kick back and run backwards. By allowing the engine time to start spinning, it helps eliminate that condition. I also wait until I see wisps of unburnt methanol coming from the headers. This is the indication that fuel is has entered the combustion chambers. I then unground the mags on both the front and rear motors and, if all goes well, they start up. Generally, the front engine that was primed with menthol starts first and the rear starts slightly later when the fuel lines fill up.

Once the crew person is sure the engines are going to continue to run (i.e. they squirted enough methanol and I opened the fuel shut-offs), then they unhook the starter and clear the starting area. Both clutches are unlocked because both engines are below the lockup speed of about 2500 rpm; the left engine is around 1000 rpm and the right is stopped. The cross box connects the output shaft of each clutch together and directs power into a transmission, which is still in neutral. Since the clutches are unlocked, the cross box and its three shafts are also at rest. I now open the throttle and rev the left side engines. That causes the left clutch to lock and the cross box to start turning, and therefore the OUTPUT shaft of the right clutch. The tractor doesn't move because the transmission is in neutral and the right side engines don't move because their clutch is still unlocked. I now press on a pedal for the right clutch. This causes the right clutch to lock. (In reality, neither clutch is solidly locked and does slip some) As soon as the other engine starts spinning, I unground the mags. They will fire up as soon as they get fuel.

And once they do, they immediately rev up to the same general speed as the left motor set. This causes the right clutch to centrifugally lock, and gently pushes the clutch pedal back up. When I feel that feedback, I know all the engines are started. At this moment in time, each engine, the cross box, and the 3 drive shafts are all spinning at 2,500 RPM, making it impossible to shift my unsynchronized transmission into gear. So, I close the throttle and watch my data recorder to see when the driveshaft speed drops to zero. This usually only takes a few seconds. I try to time it so I shift into gear as the shaft is just about to stop to make it easy for the gears to mate. If I am a little slow, sometimes the shaft stops with the transmission lined up tooth to tooth, which prevents the gear shift lever to move. In that case, very light pressure on the clutch will get the shaft to just start spinning enough to drop into gear.

At that point, the cross box is locked to the rear end and tires through the transmission and a little engine RPM gets the clutches to start dragging and propelling me forward onto the track.

It all seems complicated, but if you just think it though, it is as simple as starting the family minivan. It also requires a knowledgeable crew member because there is this tennis match type of volleying back and forth. First, they do something, then I do, then they do, and so on. For safety, it is important that each other knows the steps and what comes next. Also, the crew needs to understand what is going on, because if an engine doesn't start, I am not really in a position to see what is going on. All the gauges and switches are in a position where really only I can see them, but on the other hand, I can't see the front sides of the engines and that is really where a problem would be.

You have 4 oil pressure gauges. Why do you only have one tachometer?

Each engine has an independent oil system. Separate pan, separate oil, separate filter, and separate pump. Oil pressure is a function of engine rpm, pump quality (wear), pump design, regulator setting, oil weight, alcohol dilution of the oil, bearing clearance, and temperature. While these different parameters are likely to be the same or similar between engines, they don't have to be. Therefore, it is important to monitor the oil pressure in each engine. However, once the clutches are locked, all 4 engines MUST be turning at the same speed. There is no reason to monitor individual engine speeds because they are all the same.

The front and rear engines on a given side are mechanically and solidly coupled together, so, with the exception of breakage, they must always be turning at the same speed. The left and right pairs can be at different speeds, and it most prominently so when starting the tractor. But, I really don't care what the RPM is at idle. The only time I ever use the tach is on a particularly sandy track when I monitor engine RPM so I don't over-rev coming off the starting line, and at that time, the clutches are fully locked and the left and right side RPM's are the same.

How do you get all the engines to run at the same? That must be very difficult to get them synchronized.

It doesn't matter. Really it doesn't. Here is what I do. The idle speed is set by a small screw that holds the butterflies in the injectors open slightly. I check the idle gap on each engine with feeler gauges. This measurement is performed with the throttle cables removed. Once these are set, they really never change. Then I attach each throttle cable. The cable has a socket that snaps over a ball at the end of a bell-crank on the throttle shaft. I make sure the socket goes on essentially not touching the ball, that is, neither trying to open or close the throttle. That adjustment makes sure all 4 engines are drawing in the same amount of air. Then I watch the exhaust temperature of each engine. If the idle fuel mixture is too rich the engine runs cold and lopes really bad. This also causes a lot of raw methanol to bypass the rings and dilute the oil. On the other hand, if the mixture is too lean, the engine will run really hot and fast. So the mixture is adjusted so that each engine is lean enough not to lope, but not so lean to speed up. The idle mixture is adjusted by changing the linkage length between the barrell valve and the throttle shaft. That is about al there is to synchronization.

The only other check is done with the engines off. The throttle is opened fully to make sure the throttle plates are fully open. There have been a couple times that for various reasons, the throttle plates didn't fully open, usually because something was interfering.

Now, as fat as getting each engine to make the same horsepower, it doesn't matter. Think of it this way. You take 2 identical pickups (say 2 half-ton Silverados) and chain them to a weight. Each truck is effectively pulling about half the weight. If the weight was a pulling sled, together they would be able to pull the sled some distance, say 300', a distance farther than either truck could pull by itself. Makes sense, right?

Good. Now, lets start over and replace one of the Silverados, with a Colorado with an inline 4 engine. Both trucks still pull, the sled still moves, but now the Colorado is contributing less to the overall "system" and the result is that the sled doesn't go as far. The Silverado didn't pull its half of the sled farther than the Colorado. Both trucks started and stopped at the same time and the same place, just not as far as the first time. You still with me?

OK, now lets take our little Colorado and team it up with a 1 Ton Silverado Dually with the big V-8. Now, the little Colorado contributes way less, but both start and stop at the same time, and at the same location as the pair of Silverados did the first time.

So, if all 4 engines, make 2000 hp, then the tractor makes 8000 hp and you go so far down the track. If 3 engines make 2000, and 1 makes 1500, then the tractor makes 7500 hp and it doesn't go as far, but it still goes. The 1500 hp engine doesn't suck 500 hp out of the other three and make a 6000 hp tractor. So the important thing is to get each engine to make as much power as it can. My engines all match and each one should be relatively close to each other, but it doesn't matter. Each engine is tuned to run at whatever peak it might be.

What are all those wires?

I have two 44 channel Computech data recorders. One recorder monitors the left bank of engines and the other the right bank. I record each exhaust temperature, oil pressure, fuel pressure, and boost on each engine. I also record left and right engine RPM, drive shaft RPM, left and right tire speed, throttle position, and brake position. I tried monitoring acceleration, thinking I could calculate how high the nose was in the air, but what I found is that the tractor shakes and vibrates really bad going do the track. The vibration swamped the slight change in acceleration due to chassis angle.

When I had a single engine tractor, it was usually pretty easy to discover what the problem was. With this big tractor, it is hard to even tell which engine is bad. Sometime, going down the track, I will see something, maybe unburnt fuel coming out of a single header pipe. That would indicate a problem with one cylinder. I would swear it was left rear engine #6 cylinder. But, when I get back to shop, I discover it is really right rear engine #5. The data recorders tell you in no uncertain terms where your problem is. Sometimes, the problem is obvious. I blew an engine a while back, and windowed the block and lifted the blower. When I took it apart, I found broken rods and pistons, along with a broken cam. So, what failed first? Did it spin a rod bearing and throw broken rod pieces into the cam, causing it to break? Or, did the cam break, cause a loss of oil pressure, which caused the bearing to spin? Or did the cam catastrophically break, causes cam lobes to fall down on the rods, causing them to shatter? Regardless, the last thing to happen was a valve hung open and caused fire under the blower which caused it to lift.

I quick check of the data recorder showed a loss of exhaust gas temp on a cylinder. That is when the rod broke. But, I still had good boost and oil pressure. Then the oil pressure went away. That is when the cam broke, stopping the oil pump drive. Then a bit later, I saw the boot pressure spike way up (when the fire lit) and then drop to zero (when the blower belt and blower studs broke). So, I knew the cam and blower damage were just collateral damage and I can have confidence in replacing it with identical parts. But, when the rod broke, that is something that needs to be fixed because we don't want that to happen again. At least you know where to look.

Look at all those gauges! How do you have time to look at them going down the track?

I don't. On certain tracks, I may look at the tach during the hole shot. I check the oil pressure gauges when we start the tractor and usually when I hook to the sled. If I have oil pressure, then the engine is running. I also check them at the end of the track, so I can quickly kill the engine if it indicates a problem. But, actually driving down the track, I never look at the gauges. That is what the data recorders are for.

Do you milk cows/Are you a dairy farmer/How many head of cattle to you have on your farm?

I don't do anything with cows. I grew up in a suburb south of Dayton, OH. Both of my parents had college degrees in accounting and we lived in a 2 story house in a 1/2 acre lot in a neighborhood of hundreds of houses all built from one of a few different sets of plans. I went to college and got degrees in Electrical Engineering and Mechanical Engineering and have spent nearly all of my adult life writing software for GM automotive components, from engine controllers to radios.

My wife and I both collect cow "stuff" and have the kitchen decorated in a dairy theme. We though the white tractor with holstein spots would be eye catching a memorable. I think we were right, because you, gentile reader, remembered us long enough to find this web site and read this question.

C.O.W. Systems