Engine Engineering

The Triumph "TRactor" 4 cyl engine has some issues with crankshaft harmonics at 2 or 3 different resonant frequencies/rpms within the useful powerband. While I haven't found anyplace that says definitively that failures were caused by x, y, or z, a number of crankshafts have come up broken, often at the webs nearest the front and rear of the engine.

The TR has 3 main bearings, which would facilitate a 2nd order harmonic in the right/wrong conditions. What if the block had had 5 mains instead? That would at least change the harmonic frequencies, and if my logic doesn't fail me, it would raise the frequencies at which harmonics occur, possibly to beyond the useful powerband.

All this said, many folks chalk it up to 40+ year old cranks which were built with 60+ year old technology, so maybe a modern, custom-built crank would be a better investment than a recast 5-main block.

A better damper might be a more affordable option if you're looking to retrofit. You don't see any three main bearing engines anymore, even though they should have less rotating friction. Might be a reason!

If you added a damper to it the stock stuff might live just fine.

3 mains don't provide a great deal of support. Another thought might the block be twisting... or at least the areas around the main bearings? You might look to see if there is enough material in/around the main bearing webbings to add a strap to tie the blocks sides, into the strength of the mains........

what I'm talking about, but on a Toyota block

You could also change the weight of the components to change the frequencies. Lighter or heavier flywheel / damper / clutch assembly could do it.

It's often the power pulses that cause these vibrations, and making more power only makes them worse because the pulses get stronger. The damper is specifically designed to absorb these pulses.

If you replace the factory damper on a Miata engine with a billet "underdrive pulley" and you make more than about 150 rwhp, your oil pump won't last until the next oil change.

As I said, there are a number of things that can cause or accentuate the natural frequencies. Harmonic balancer or lack thereof, out-of-balance crank/flywheel/clutch/driveshaft just to name a few. Weights and locations of the rotating components in the driveline would also affect the harmonic frequencies and severity of the harmonics.

Hadn't thought about deformation of the block. That's an interesting idea, and not one I've seen in connection with the Triumph engine. Adjusting component weights to change harmonic frequencies would be an interesting study... wonder how accurate you could get on paper (without doing practical experiments and destroying a pile of cranks).

I put an electric fan in mine six weeks ago, which required removing the stock fan and the heavy cast hub it mounts to. The hub is said to act as a dampener of sorts, though it doesn't strike me as a precision-balanced component. Symmetric, but rough cast (rather than machined). This got me to scheming about the broken crank issue and possible solutions through better engineering.

Seems like, for the sake of being thorough and money-no-object, one would cast a 5-main block, use a custom-made crank to accommodate the additional mains, and a retro-fitted harmonic balancer.

Not that I have money, just thinking aloud.

Google search: Moldex crankshaft Triumph

If it's a budget build and the intent is to not break crankshafts, then the usual race-use inspection and prep on a stock crank should always be the prime path. Although that's true for most any OEM part.

If you got a non-cross flow Ford kent head and a later Ford Kent block with 5 bearings for the crank, it could look similar to a triumph engine.

A lot of the engineering in my industry is not creating new products but finding the right existing product to do the job. I realize this isn't your intent however, but it should be a consideration; an engine swap with characteristics that you desire.

1) Install an effective crank damper
2) Replace the rods with something lighter (they are >> 800g).
3) Lightened flywheel.
4) Balance all rotating components
5) Blueprint the drivetrain to minimize backlash (the motor doen't run in a vacuum) it's connected to a clutch, gearbox input spline, cluster gear, 4th synchro, output shaft, 2 U Joints, diff input pinion, diff side gears, diff planet gears, axle shafts, and finally the tires. Each mechanical connection has backlash .....

Edit: 6) Install Megasquirt, Megajolt, etc, to precisely control spark and fuel,
Knocking you can hear, but pinging often cannot; both will do a number on your crank. That sloppy mechanical gear driven dizzy is well known for spark timing scatter, even when brand new. (How old is the one in your car?)

In reply to JohnInKansas:

I worked for Tremec transmissions in Mexico for awhile. One of my jobs was to eliminate frequencies in the transmissions that that would appear in teh normal operating range of the transmissions. These frequencies caused NVH issues as well as premature failures.

I modeled the entire tranny in ANSYS and used the software's frequency analysis. I then took those numbers and did some real world testing. I put the transmission on a shaker machine with LOTS of accelerometers and compared those values to what I got out of the ANSYS model. THis allowed me to verify teh ANSYS model.

I then used the model to increase or decrease size and shape of the individual components and reran teh test on the model. Once all of the frequencies were out of the normal operating range, we made a "one-off" transmission. Sure enough, it was MUCH smoother, quieter, and lasted longer in testing.

The changes were made on a larger scale for production.

I guess what I am saying is: If you are good at modeling and have access to ANSYS or one of the other programs, you could make a virtual model and try to design these things out.

Some of the frequencies were as easy as changing weights by 2 %.

Rob R.

I saw an interesting article/book online about the engineering history of a Rolls-Royce (maybe Betley) inline engine, starting from the teens and working through to the end of production in the 50s/60s. Every time they tried to make more power, either through more compression or higher RPM, they had to deal with harmonic problems in the crank. These problems would manifest elsewhere too, like breaking bellhousings.

Another thing to look into are the Ford small sixes. They went back and forth between four and even main bearings.

In reply to erohslc:

I'm running a brand-spanking-new Pertronix Flamethrower magnetic distributor with matching coil. I don't really want to convert to fuel injection at this point, though I might in the future. Megajolt would have been cheaper than the Pertronix setup, and sounds like it would be pretty friendly to use.

In reply to wvumtnbkr:

I've got access to Solidworks 2012... I'll have to look into what it can do in terms of frequency analysis. Quite a bit of modeling involved in an engine, but more in a tranny.

In reply to Knurled:

That's a fair point too. Make a change to reduce the harmonic of the crank and you may see it elsewhere in the form of other broken parts.

MG started the B series motor on 3 mains when it was 1500/1600cc in the MGA, the early MGB 1800cc motors wre also 3 main. They then went through the laborious conversion to 5 mains. Given how cheap and resistant to change the Brits were, there had to be an overwhelming engineering reason to do that.

Triumph made the Spitfire motor in sizes from 803cc to 1500cc, all were 3 main. I have never seen an 1147 or 1298 break a crank, even in the small journal versions (although they were hard on main bearings) but I have seen several busted 1500 cranks and every one broke between the center main and #3 cylinder. I guess the crank 'whipped' or flexed bad right there, causing the breakage and since the FO is 1 3 4 2 that sort of makes sense.

Funny thing: I have never seen a 4 main GT6 or TR6 motor break a (much longer than a Spitfire) crank, even when flogged pretty hard. The firing order is 1 5 3 6 2 4, which does put the firing impulses in roughly the same spot as the Spitfire. But it seems that instead of breaking cranks they shake the flywheels off. http://www.expeditionlandrover.info/2trSite/Net_TR6clutch.htm

Must be something to all the harmonics you mention.

Curmudgeon wrote: Triumph made the Spitfire motor in sizes from 803cc to 1500cc, all were 3 main. I have never seen an 1147 or 1298 break a crank, even in the small journal versions (although they were hard on main bearings) but I have seen several busted 1500 cranks and every one broke between the center main and #3 cylinder. I guess the crank 'whipped' or flexed bad right there, causing the breakage and since the FO is 1 3 4 2 that sort of makes sense.

A fellow I knew in Florence, SC used to work on "foreign cars". He would swear that Triumph had done something different with the 1500 cranks. He said one would be fine and the next a POS. He thought they had screwed around with the cast iron formula they used for the cranks and had cheapened it too much. I've heard similar from others. British Leyland was definitely known for being penny wise and pound foolish.

One theory I heard was the 1500 block flexed under load and caused the crank breakage. I dunno; on the the last broken one I saw (1500 Midget) the main bearings looked perfect.

FWIW, I'm working with a 2238.