An Entertaining & Informative Vintage Automobile Internet Magazine

Part II: Bending the Rules in NHRA Championship Racing Using a Roller Bearing Crankshaft

By David Greenlees: This is the second part of this story and this installment of it features a custom built-up-ten-piece roller bearing crankshaft for a Mark IV 427 c.i. Chevrolet V-8 engine that was used in NHRA Championship Race meets in the 1970s. If you are new to this series of posts you can look back to Part I where we covered the background information about how and why these crankshafts were manufactured.

This particular crankshaft is a more exotic and costly proposition than the one-piece crank in “Part I.” This shaft was precision machined from individual pieces of forged aircraft alloy steel that was heat-treated and surface hardened to prevent wear by the bearing rollers.

  • Ten-piece-built-up Mark IV 427 c.i. Chevrolet V-8 roller bearing crankshaft (top), five main bearings and roller thrust bearings and side plates (middle), eight connecting rod bearings and thrust washers (bottom.)

Each individual section of the assembly is precisely located to its adjoining half by the use of hardened steel dowel pins that fit into eight bored and reamed holes in each piece. The two sections are then held together by a fastener that passes through the center of the bearing journal.

By using this form of construction the connecting rods (not pictured) can have a full one piece round lower end and are not be split and held together by bolts (a weak point) as is used with a conventional style rod. Each double row roller main bearing is placed on one half of its bearing journal and the two pieces are assembled and then fastened together; this assembly method is also used for the connecting rods and bearings.

As stated earlier in “Part I”:

“We have decided to not name names and companies involved with making this equipment and the individuals using this equipment out of respect to their surviving family members and because of possible litigation.”

“Because of the decision before anyone starts crying foul and says this is a hoax or the photos were created in Photoshop, rest assured that what you see in the original small period photos (2″x 3″), which are in our own collection are real. The only thing that has been removed from the photos we are sharing with you is the logo of the engineering company that masterminded and built these lower end assemblies.”

“Also in our possession is a custom made forged V-8 crankshaft (one of three made and the only one that survived,) a set of main and connecting rod roller bearings, and one connecting rod that was used in a 427 c.i. Ford engine during the 1960s in Nascar Grand National Series competition that resulted in winning the championship laurels.”

28 responses to “Part II: Bending the Rules in NHRA Championship Racing Using a Roller Bearing Crankshaft

  1. Thank you for sharing this, David. You are correct that the crank was costly. I would only add that it was probably VERY costly. I am going to guess that they assembled the crank to grind the journals then disassembled it to add the rods. Do you know if the main bearings were split or, I am guessing again, solid hardened cylinders that were held in place by a traditional caps?

    The old single-cylinder motorcycle engines from the 50’s had roller bearings on the rod. The pin for the rod bearing was pressed into the cylindrical cheeks of the crankshaft.

    • I had a JAP 1924 Speedway engine that big end has its end conic held in place by a flat nut so where the two end shafts, and saw a 2 stroke twin Gutbrod engine from the 50s tha was assembled the same way with cones and nuts .

    • It wasn’t just the 50s bikes with that – the Suzuki GS1000 (and doubtless others) from the 70s/80s had roller bearing cranks, too.
      At a time when motorcycle engines were often considered fragile, these cranks were expected to last 200000 miles or more, and were often used in drag racing.

      • Kawasaki used roller bearung crankshafts in the KZ900-1000 bikes.
        The main trouble with a roller bearing crank is they don’t hold oil pressure, so you need 2 oiling systems, a high volume pump for the crank, and a high pressure one for everything else.

  2. Bugatti used a similar construction, and there were the Hirth crankshafts used in many engines, including IIRC the DB 601 used in (among others) the Me 109.
    Quite a lot of single and V-twin cycle engines had full roller crankshafts, but I never saw one with a pressed in crankpin; all I have worked on had the crankpin fitted with a taper and keys, held together by a nut at eack end of the crankpin. Which isn’t to say none were pressed.

    • XR 750 Harley 1972 on had a straight pressed in pin then welded in place. Vincent had a straight pressed in pin as well. Caged rollers as well like all Harleys

  3. Hirth supplied multi-section roller-bearing crankshafts to Porsche, and my ’52 356 1500S had one. They were fine at higher rpms but did not like being lugged at low rpms. With the crank and road assembly out of the engine, the tops of the connecting rods could be moved several degrees sideways.

  4. When you ran the first article of this series I wondered, if these bearings cut down on friction so much, why were they never developed for passenger car use? Yes, I realize it would be expensive, but the bearings would be under a lot less stress in a family sedan. They may be able to cut some corners on hardening and such.

    After all, there was a time when fuel injection, four-valves-per-cylinder, and variable valve timing were only found in racing motors. Now I’ve got all three in what is essentially a family station wagon.

  5. Sounds like the start of Dragnet

    “The story you are about to see is true.The names have been changed to protect the innocent”

  6. Reminds me of the good old days of street racing in Springfield Mass in the late seventies. We had one class of cars for drag racing in the North End. The class was called……
    “Run what you brung!”

    We were running high 9 second quarters…..not too bad for high school kids!

  7. For what it’s worth, I worked on a small team developing a roller element bottom end on a V-4 Stirling Engine in the late 80’s. It was a low HP (75HP) multifuel engine .5 litre swept volume that would run on 4 different fuels. The project was some what successful but unfortunately it ran out of money before everything was brought to fruition. It went on the shelf at NASA Glen in Cleveland, OH and remains there to this day. The roller element aspect of the engine was sound after going through several iterations of crank hardness to obtain reliability.

  8. The first version made sense to me this one however split roller cranks have big problems with twist. It would also require split big ends on the conrods otherwise it would be impossible to install in a Chev Big Block. Today and even back then you could take 1/4 ~3/8″ diameter of the journals and achieve virtually the same thing. A full roller engine normally only needs enough pressure to lubricate say 10 psi at most. Seems like an exercise in skirting some badly written rules rather than cheating to me. Really the only power gain is from the down graded oil pump and that isn’t that much due to the much higher volume required

    • In the time before “modern improvements” with plain bearings and motor oil were still years down the road a correctly engineered and constructed roller bearing set up like this was still capable of producing more horsepower than a conventional lower end. They actually do not “twist” anymore than than one piece cranks. Both ball and roller bearings used in this way do not need a higher volume of oil – in actual use the amount of oil needed is much only 20% of what in normally used with plain bearings. This crankshaft was assembled with one piece rods and raced so it is actually possible to put it together.

      • Wow that’s a surprise pistons installed from the bottom on a 427. That would be a headache. Or perhaps assemble the crank in the journals.
        I have no experience in bolt together cranks but do with press together keyed / unkeyed units and when the hp goes up they do twist. On my GSX1150 I gave up in the end and just welded it to prevent the twist. That engine however did suffer from boost and nitro.
        Thanks Dave

  9. As I remember, Vincent singles and twins used roller bearing lower ends. The talk back in the early 50’s was that at some point the rollers would slide rather than roll. Some JAP engines used them also.

  10. Certainly enjoyed both volumes on the use of roller bearing bottom ends by the folks who wanted more output from their big block V-8’s. Although expensive to do there was one car company that used roller bottom ends for a good number of years – Panhard of France in 1945 started to use it for their flat air cooled twin engine. Stayed in production till the end in 1967. One unique feature was the use of smaller inverter rollers between the load bearing rollers that turned in the opposite direction….even less friction than with a cage. The engine had other unique features such as torsion-bar valve springs that were linked at their free ends which increased the closed pressure on the shut valve when the other was open.

  11. I read all the assumptions here of “High Volume/Low Pressure Oil Pumps Needed” and all these wild ideas, You do realize that this is essentially how ALL 2-Cycle Snowmobile Engines are made, and they are “Dribbling” oil into the Crank Main Bearings if oil injected (Which some are Ball, but Roller has been used) but most race engines are “Pre-Mix” and the Con-Rod on either style ONLY gets, Oil Diluted with Gasoline “Mist Splash Lubrication” and turn easily 10000 RPM and easily make 200+ HP per Liter naturally aspirated… As David Greenlees pointed out, this engine needed very little of any kind of oil pump for this type of crankshaft.
    Also someone stated that inserting/assembling the pistons from the bottom side would be difficult, not really, you put the piston/rod in from the top all assembled, then because the crank is broken in so many pieces it becomes almost “Fluid” in that the rod can be slid over the crank pin and the bearing cage inserted between… OR you can assemble the piston rings only slide the piston down the bore until the Wrist pin bore pops out the bottom of the bore with the piston turned about 45 degrees, insert the wrist pin/bushing/bearings, clips and then turn back 45 degrees and push assembly up the bore, assemble the crank section…but that still requires the usage of the crank “Fluidity” that is afforded by the fact its in ten sections, the nice thing is they are not necessarily “pressed” together, but using 8 floating pins (that are made of some very good material) and 1 connection “fastener” per section, this system is still very strong. Unlike pressed together cranks that basically have TWO narrow joining surfaces as locating “pins” this has NINE joining surfaces per “split” so it is very stable.

Leave a Reply to Matt Cancel reply

Your email address will not be published. Required fields are marked *