65% balance factor test

Made a jig with a tiny brass tube superglued & taped to a heavy block, set it on a jack stand next to the Norton 750 motor. Inserted a 1/32 wire sharpened to a needle point. Taped a piece of sheetmetal to the timing cover. My motor is now balanced at 65% wet (72% dry) in solid featherbed frame. Sat on the bike with brakes on & Revved it to 6000 while my wife Beth pushed the sharp wire against the sheet metal. Put the sheet metal under a microscope. (photo of simple scratch tool below).



vibe scratch test circles are round at both 4000 and 6000 RPM but slightly elongated front to rear at 6000. This means that the Balance factor is just right and certainly not to low (see photo below). These drawings of what I saw under the microscope - not actual photos. Note that these scratch circles are only about .010" in diameter because of the lightweight pistons. They would be much larger with heavier pistons.



It could probably go down to 62% But I'm going to call it good.

When the motor stakes too much front and rear from too high a BF its because the flywheel counterweight is shaking the motor to the front and rear more than up and down.

My balance factor used to be up at 72% dry (80% dry) and then scratch circles were too elongated front to rear - indicating too high a balance factor (see photo below).



You can do you own tests but I expect you are going to get the same results regardless of frame or rubber mounts. Remember that a rubber mounted frame has the swing arm and rear wheel and a bunch of other weight attached to it and the crank is shaking against that weight.

Fascinating tracings and reasoning. implies to me the G forces are pretty much spread around averaged out inside engine and out to frame and tire patch, about as good as physics allows. Your conversational bargaining skills to get wife away from her routine is even more impressive. ah, is this a wet or dry BF reporting?

Can it show anything other than loops and whirls ?
Be interesting to try it with a different balance factor, wouldn't it ?

And, the big test, whats it like out on the road.....

This test is only valid for the long Carrillo rods and lightweight pistons. Standard rods and pistons would act differently, just so everyone does not run out and throw a 65% balance factor into their Norton tonight.

Anything that changes the reciprocating weight or the rod-stroke ratio is going to have it's own optimal balance factor. And of course the frame makes a difference too.

When Jim Schmidt ran standard rods and pistons he had different recommendations for everyone in the manual he published, 65% was for low rpm street touring, and the racer got 85% to make it rideable.

Norton 88, 99, 650, 750 and 850 engines with stock components all have either different reciprocating weights or rod-stroke ratios, and each would need unique attention to the balance factor.

No doubt the longer the rods and lighter the pistons and lower the rpm, the lower balance factor you can go in a rigid frame. Also a billet crank or some other custom crank is going to change a lot of things, as will the overall flywheel weight change the rpm or frequency a parallel twin and it's jump-rope crankshaft vibrates at.

My old race manual specs were mainly based on hear say and I was able to reduce vibes by lowering the piston weight and that threw off my judgement. So I no longer agree with the 85% BF. Higher balance factors do not work better at higher RPM. I used to think so, but no more.

Rod stroke ratios do not make as much diff as you would expect. For instance - at top & bottom dead center it makes no diff at all. And the shorter rods increase the severity of the vibes regardless of BF.

What makes the most difference is the weight of the rod cap and how that effects the apparent weight of the small end. For instance - a 10 pound rod cap would reduce the apparent small end weight of the rod (when weighed horizontally). The same rod with a one ounce rod cap would seem to weigh heavier on the small end when in fact the small end weighs and shakes the same with either rod cap. It gets complicated & confusing.

The accurate way to balance would be without the rod cap, but thats impractical.

jseng1 said:

My old race manual specs were mainly based on hear say

Click to expand...

Well the old manuals of yours I have from the 80s/90s talk about first-hand experiences with various balance factors. I will dig one out and get some quotes..

If you are retracting the advice in your old manuals which you now say were based on engines with lightened pistons, and your current opinions are based on your experiments with your lightweight rods and pistons, then that means you actually have no valid advice on balance factors for those who's engines are running non-lightened cranks, rods and pistons ? Or have you done that research too?

Probably only a small fraction of 1% of Norton twins are running lightened components, or raced for that matter, so information and research about balancing twins with stock components would be at least 99 times more useful to the general public.

As far as reputable advice from top tuners; T.C. Christenson's Norton dealer/tuner John Gregory said that he liked a 78% factor for all his engines regardless of their use or spec. R.J. Reynolds, who is known in Norton tuning circles has run very high balance factors on Norton racers, he even talked of one with 100% once, among other unusual balance tricks like primary chain tensioners and external balance weights. Ron Wood used a low balance factor in his dirt trackers that had solid billet cranks, which is not really applicable to stock-parts rebuilds. 650 Nortons had a lower balance factor than the 750 Atlas. Some people I know who ride Atlases like the factory balance job and some don't.

I am not a Norton tuner, but the best thing I have found first hand that works well for adults riding a street bike is to tune and balance it for low and mid-range rpm and power, as they mostly were from the factory, and gear the bike so when you are on the highway it is turning low rpms, makes for a really enjoyable featherbed roadster that will still cruise at highway or even illegal speeds, and it will also see the engine lasting longer.

Running stock parts combinations at high rpm, over 6000, is going to use them up quickly no matter what you do with the balance factor.

Of course if you are that 1% that can afford to be able to use up lots of stock parts, or to buy a $6000.00 set of replica crankcases, crank, rods and pistons, then all you have to do is follow their manufacturer's advice.

The whole point of this post is to show everyone that there is a way to determine the best BF no matter what parts you are using. TC Christian liked high balance factors in all frames, Ron Wood liked low balance factors. Norton's from the factory came high or low depending on the year and model. BFs range from 85% to 52%. What this indicates is that everyone has been guessing or copying, (myself included), and it has all been subjective. I started to suspect numbers that I heard about and also the numbers that I used to push in my race manual. This is not a perfected art and I and everyone else has made mistakes.

What I hear and see is a lot of regurgitated advice from experts who don't agree and can't all be correct and different at the same time. I also get a lot of opinions and verbage from people who have never made a test of any kind other than possibly - seat of the pants.

So I made some crude test equipment and got some concrete factual answers.

Now lets see someone make a test of their own. Look at the tools I've shown above. Someone please make the test or similar test (but accurate) on a stock Commando. Refine them or improve on them anyway you want. The tools above take repeated attempts and practice to be successful - you need a microscope and you have to be able to make adjustments and figure out corrections if the test doesn't work the first few tries.

Best yet - have a removable plug (Nourish crank style) so you can change the BF. And try a couple of different & widely spaced RPMs.

The test results would be some valuable info.

Many factors on BF to keep in mind. One[1], the best-least-averaged forces on outr 360 type cranks and bearings and cases. Two[2], the best-least-averaged forces transmitted to the frame and/or pilot. Three[3], BSA research showed best-least-averaged sling forces obtained with 52% BF, period end of story no matter what your beliefs or emotional state is. Four[4], each bike and rider may feel better with a non circular sling force of higher BF. With lighter reciprocating items and/or longer rods the magnitude of the crank sling forces is lowered so BF less important to feel and crank forces to resist over a wide range of BF. Isolastics allowed Norton the idealistic 52-54 BF to spend less on less mass of metal materials of heavier counterbalanced cranks than non iso models. Five[5] rubber mounted Cdo's can get away with more perfect internal forces for less frame and pilot vibration. Six[6] Jim's engine is solid mounted so the frame mass is also added to the engine orbital equation of what BF allows him more circular traces.

The mystery then is why prior British factories and private racers that had decades of practice to come up with their least expensive most effective BF they could still sell that bikes and pilots could tolerate, a reasonable time. Someday when I'm richer will send around a device to clip on bars or frame to objectivity measure the various planes and magnitudes of vibes.

Jim, don't be to quick to poo poo the "seat of the pants" method of determining balance factor on our 360* parallel twins, after all, isn't that actually the final test...is this thing rideable at the rpm we wish to run? The BF is dependent on this seat of the pants test, Triumph put a bunch of riders on bikes with various BF's and then let a poll decide what factor was most enjoyable for the bikes intended purpose, what could be better? What is any scientific test going to do for us, I look at the circles and elipses and say 'ok, so what?' The engine may be happiest, bearing load wise at 52% BF but unless that thing is in an isolastic frame then that low of a BF is going to have a high vertical vibe component that us humans are sensitive to and probably would be unbearable in a solid mount frame. So we go to a higher factor to make those little circles into elipses moving the vertical shaking into a more fore and aft shaking which we can tolerate, it's all a really big compromise depending on engine design/tune, frame design/materials and can't be made perfect or even close without a different engine design configuration or balance systems such as balance shafts, dummy rod and piston wt etc. We can move the vibes around until we find a BF that our bods can live with, racers will want the same engine smoother at a lot higher rpm so will tolerate the vibration at an rpm that street riders wouldn't.

It would be great to be able to say, Here is the perfect BF for this engine, in this configuration with this frame, these accessories at this rpm. I think the factories would come up with a best compromise dependent on their testing with the bike configured in the form they were selling it as and run in the manner tested for...street riding right? Anything changes and all bets are off. Sloppy eng. mt. bolt holes, clamped on accessories, different state of eng. tune, all these things and more can change where an engine vibrates and whether it's in a tolerable rpm range for it's intended purpose.

I guess what I'm saying is if there is one thing I've learned about BF's is that all they do is change the plane our 360*'s vibrate in, they still vibrate but hopefully we've picked a factor that works for us at the rpm we will ride or race at, it won't be as good elsewhere but that's the compromise. So, as long as people understand that the BF you come up with for Jim's bike is a good starting point for a similarly built machine, then all's good, there is no 'one' right answer here that works in all circumstances. I applaud you Jim, for your work on this BF subject but your biggest contribution to the BF in Nortons is your work to reduce the wt. and improve the geometry of the rods/piston assemblies...


John Healey, Britbike forum Triumph moderator, once commented in the Vintage Bike mag that they were working on a pre-unit Triumph 650 with duplex frame that vibrated unbearably at highway speed, that is until they bolted back on the highway pegs that had a bar that spanned across the front of the engine, clamping those duplex frame down tubes together. After that at cruise it was as smooth as a Goldwing, (he may not have said Goldwing but you get the idea). The balance factor allowed the engine to vibrate at a frequency (at cruise rpm) that matched that of those frame tubes and everything including the rider vibrated together...yikes, not good but you can see how one thing changed the formulae. Bolt the highway pegs/bar back on and it changed the rpm at which those tubes would vibrate, to an rpm not sympathetic with the engine r's.

I've read about people with similar set ups to my Triumph project and the recommended BF's have been anywhere from 65% to 85%, each one claiming to be smoothest at "X" rpm while the seemingly similar bike claimed "not good" at "X" rpm...what to do? Jim's taken the first most important step and that is to reduce the reciprocating wt., as the less recip. wt. we have to counterbalance the less out of balance our crank is when approaching/leaving TDC and BDC. Have a professional dynamic balance job done and take a poll on BF and see what your 'seat of the pants' tells you...Mark

Ok Mark I can agree with much of that but even so - using a vibe scratch tool will still give you a much better indication of what vibrates less & which direction - you can bolt it on any style frame or the motor near the crank end. It will show you how much vertical shaking is happening at riding on any RPM. And it will show if higher BFs are best for hi RRMs or if that assumption is just bogus.

Note that your arms & hands are not just pressing vertically on the handlebars but leaning on them more horizontally - especially if you're on a cafe or race bike. The horizontal vibration plane is just as important as the vertical plane. Frame resonance will only happen at a very narrow frequency and all you have to do is change the RPM a tad.

And who has proven what BF is best for the crank & bearings with a balance of vertical/horizontal vibration? What is this BF and what is it based on? Just math or evidence we can look at?

As for seat of the pants - of course we ride the bikes. My Atlas may be smooth and comfortable but its just a description. I'll say that is seems to me smoothest at 4000RPM. Below that doesn't matter because there just isn't enough vibs to be concerned with. 6000 to 7000 RPM is still a buzz but nothing is breaking and parts no longer crack or fall off the bike as they used to.

I feel like I've found a way to find the best BF for these motors. But that doesn't mean anyone will adapt it or that its going to change established thinking.

'The engine may be happiest, bearing load wise at 52% BF ' If you believe that is true, take your standard commando out and rev it often to 8000 RPM. You will find out the truth as many others have before you. My 850 motor is balanced a 72% dry, and you really wouldn't enjoy the experience of riding the bike in slow traffic. My feeling is that if I owned a stock 850 commando, I would balance it at about 65% dry, and I would secure the ends of the swing arm pivot to the Z plates with flexible but very restrictive mounts , if that is practicable. My Seeley has the motor rigidly mounted, and is not uncomfortable, however you would not want it as a road bike unless you were only going to challenge the local louts to speed duels on Sundays. A commando set up as I've described, would not be unbearable. When the handling of the bike depends on the head steady on the motor, that is idiocy. I really wouldn't ride a bike like that, it is asking for a skin rash.
The 52% balance factor is obviously used to stop the bike from rocking backwards and forwards while it is idling and being ridden at low speeds. Norton had obviously seen the CB750 and the 60s two strokes running, and were trying to compete with a very old design. The trouble is that Force = Mass X Acceleration, and when you spin an out of balance crank at high revs ...... ? So you have to choose how you want to used your bike. Norton were almost successful - for what it is, the Commando is not too bad.

I recently repaired my spare set of crankcases by welding a wide close fitting 8mm aluminium plate around the drive side bearing. I believe that as it cooled after welding the plate would have shrunk and compressed the crack which ran horizontally through the bearing housing. Has anyone else here had experience with this type of repair ?

acotrel said:

'The engine may be happiest, bearing load wise at 52% BF ' If you believe that is true, take your standard commando out and rev it often to 8000 RPM. You will find out the truth as many others have before you. My 850 motor is balanced a 72% dry, and you really wouldn't enjoy the experience of riding the bike in slow traffic. My feeling is that if I owned a stock 850 commando, I would balance it at about 65% dry, and I would secure the ends of the swing arm pivot to the Z plates with flexible but very restrictive mounts , if that is practicable. My Seeley has the motor rigidly mounted, and is not uncomfortable, however you would not want it as a road bike unless you were only going to challenge the local louts to speed duels on Sundays. A commando set up as I've described, would not be unbearable. When the handling of the bike depends on the head steady on the motor, that is idiocy. I really wouldn't ride a bike like that, it is asking for a skin rash.
The 52% balance factor is obviously used to stop the bike from rocking backwards and forwards while it is idling and being ridden at low speeds. Norton had obviously seen the CB750 and the 60s two strokes running, and were trying to compete with a very old design. The trouble is that Force = Mass X Acceleration, and when you spin an out of balance crank at high revs ...... ? So you have to choose how you want to used your bike. Norton were almost successful - for what it is, the Commando is not too bad.

Click to expand...

There is nothing in any engineering textbooks about the balance factor and bearing life ??
If the bearings are strong enough to survive a ~25 lb weight whirling around at 6 or 7 or 8 thousand revs per minute, with what HAS to be an unweighted/unbalanced flywheel to accomodate the reciprocating portion of the equation, then they essentially don't care what the balance factor is going to be. Combats simply didn't have a strong enough bearing specified for the job and revs available, and once that was rectified, few bearing problems reappeared.

The Commando, and the Atlas before it came out WELL before the Honda 4 or Z1 Kawasaki, so they certainly didn't get any lessons from those.
The Commando balance factor chosen was for comfortable cruising speeds - if you have ever seen a stock Commando at idle, the motor joggles around, quite a bit, depending on how carefully you have set the idle, and smooths out with a few revs on board. If the iso's are shimmed to factory specs, the rear should play no part in the steering.

You keep repeating the same mis-truths about even basic Commando facts of life ??
Perhaps you have never ridden one ?
Peter Williams lapped the IoM faster than a Manx had ever done, on essentially a slightly souped up Commando.
The later JPN racers got more elaborate, and in pursuit of a real race performance, not a souped up road bike...

Long rods will help make an engine smoother, that is a fact and here is why:

If you had infinitely long rods, then the acceleration of your piston between top dead center and bottom dead center would match the acceleration of the mass of the counterweights in the crankshaft and they would oppose each other like the pistons in a boxer engine. At 100% balance factor in the infinite rod engine the engine would move little in the up and down plane, but a lot forwards and backwards.

You could also balance your infinite rod engine to 0% and have almost no forwards and backwards but a lot of up and down.

When you try a 50% balance factor in your infinity rod engine, you are going to start making circles. Because you no longer have your perfect counterbalance to the reciprocating pistons, you are going to start going up and down but not forwards and backwards as much. You notice that the higher rpm you operate the engine at the larger the circles become. The engine is orbiting it's crankshaft center-line.

Now you have to make the engine fit into a vehicle so you have to cut it's height down so it fits, and you end up with rods about 6" long. You keep the 50% balance factor but you notice that although the engine still does about the same thing at low rpms, the higher the rpms go it's orbit around it's crankshaft goes elliptical. This is because with finite length connecting rods, the acceleration of the reciprocating pistons no longer matches that of the opposing crankshaft counterweights at from TDC to BDC. While the position of the crank counterweight is at the same acceleration at any degree of travel as it was in the infinite rod engine, the reciprocating assembly has a completely different acceleration at any given degree and is even different on the way up than it is on the way down.

Because of the finite length connecting rod, the acceleration of the crank counterweight only matches the acceleration of the pistons at one rpm measurement, at all other rpms there is a compromise.

At very low rpms the out of balance of the engine does not have enough energy to distort the engines orbit into an ellipse, but the higher the rpm the more energy the out of balance of the engine has and the orbit will be squashed. At 0% balance the circular orbit will turn into a vertical ellipse, at 100% the circular orbit will distort into a horizontal ellipse.

If the flex of a chassis lets the engine move in the plane it wants to go at high rpm then it will feel smoother than if it was stiff in that plane.

In the end there is no way to balance a single or parallel twin engine, you can only try to match the way it moves to the type of chassis it is in by changing the chassis or the direction of it's engine's largest movement.

If the chassis is a large spring that is flexing from the input of the engine, then hopefully the rider is holding onto part of the spring that is not moving as much as the part the engine is bolted to. If the chassis were perfectly stiff moving exactly with the engine, then there would be no relief from vibration unless the chassis could be made infinitely heavy, or the engine could be made to weigh nothing at all.

That's a well put description, Ben. I don't think you could make it any clearer without including the relevant equations and a bunch of illustrations. This is certainly an issue we've beat to death on this forum already, but I think your description is one of the best so far. Anyone who wants to understand the issues in more depth can find the relevant discussion and equations in any one of a number of good automotive engineering books.

Ken

Yeah Beng - that's very well put. I could have used your info years ago and it would have saved a lot of time. But I would still have made the scratch test to confirm and get real numbers that I can record.

BTW the scratch tool should employ thicker stiffer precisely fitting wire so it doesn't flex. I had to move mine up close to the primary cover so it only extended a max of 1/8".

Going to longer rods helped bring town the vibes but I have no way of knowing how much - the longer rod & short pistons was a mutually benefiting solution. I'm not sure about BFs in the 50s I would have to see some evidence before I would try such a low figure. I did try 49% in a solid frame once and I thought it was no good.

In electric digital age 2 other recording options occur to me. Tiny ink jet head aimed at paper or needle emited spark gap through paper. The pulsed timing of the marks would also reveal the speeds of the various parts of the orbital.

How about a load cell and digitizer with a memory stick, mounted in place of the front isolastic mount ?

Yes gotta be a fairly simple digital phone age way to record off bar mount or motor mount. Ms Peel should be able to embarrass Goldwings for smoothness sense so need objective ways to find pecking order. Only reason not to rubber mount and be done with vibes is the poor handling yet Peel cheap links totally solved vibes and handling to point its no longer anything I think about. i'm now interested in the orbital shape and phase only as it affects the power thrusts from rear patch as various angles of attack. Peel so completely disappeared that the only thing getting through was the rear patch harmonics so allowed me to play tunes within the max tire traction zones up to 10% slippage, which gets into the digital nature of physics and felt as series of bumps in the rear of stronger accelerations. By accident Peel's crank has feature to change BF in the cases.

If anyone can find the BSA reasch article on crank BF vs bearing-case loads please post to me.

I don't have a problem with the low BF used in the standard commando, however I'm certain it is done along with the isolastics to provide smoothness at low speeds. I suggest if you want to make a commando into a fast bike, you have to choose how you are going to set it up. The crank which gives smoothness at low speeds is often horrible when revved hard . My worry is always the BIG BANG. - I suppose Rohan is correct, rider comfort is the issue - it is very uncomfortable when that happens. Incidentally, there have been some comments that a high BF suits a rigidly mounted motor better. I suggest the isolastics would be more compatible with a high BF than a rigid mount. The rigid mount and high balance factor is well proven in old style race bikes. If you read that article on Rex McCandless and the featherbed frame, you will see the emphasis on rigidity between the steering head and the rear tyre contact patch. Also rubber mounted motors in race bikes always seem to give a less powerful result than when rigidly mounted. My feeling is that rather than rubber mount the engine, a better approach is to build spring into the frame, and avoid stress raisers.