Balancing IC engine

[John ViettiParticipant @johnvietti39406]

I'm building the Kittiwake and made the crank and rod to the plans including the crank counterweights. It looks massively overbalanced.

What is considered the best way to balance the engine? I'm aware of the concept of adding a weighted blob on the crank, the blob equal in weight to 1/2 the weight of half the conrod measured at both ends. The counterweights are then whittled down till the assembly balances on knife edges. ' Ive used this method on hit and miss engines but they are obviously made to run much more slowly.

Is there a better way or is this good enough?

Thanks for your input, John

[John ViettiParticipant @johnvietti39406]

[HopperParticipant @hopper]

It's a bit more complicated than that. On full sized motorcycle engines the balance weight is equal to all of the rotating mass plus 50 to 60 percent of the reciprocating mass.

Rotating mass includes the crankpin, big end bearings, retaining nuts and lock tags, so everything that goes around in a circle. Plus the weight of the large end of the con rod.

The reciprocating weight is regarded as the piston, rings, gudgeon, cirlips and the weight of the small end of the rod.

Once you have calculated your weight, you make a disc of that weight and clamp it where the crankpin would go. (or deduct the weight of the crankpin and clamp the dummy weight around the crankpin.)

Then you add or reduce your flywheel's permanent counterweights to get it balanced with the dummy weight.

I would assume the same would hold true for model sized engines. There is endless argument in motorcycle circles on which percentage is the best balance factor. 50 per cent was the traditional factor. But now some prefer 55 or 58 or 59 or 60. It's a matter of religion, like synthetic oil and Chinese lathes.

Edited By Hopper on 02/05/2020 03:56:49

[Roderick JenkinsParticipant @roderickjenkins93242]

As Hopper says. For a single and in the absence of a flywheel then all the balance weight is in the web opposite the crank pin. A simple single like the Kittiwake cannot be perfectly balanced but 50% of the reciprocating weight seems to be the consensus. Since Westbury designed the Kittiwake to be a performance engine for a hydroplane I suspect he probably got it balanced as best he could.

Stay well,

Rod

[HopperParticipant @hopper]

Not sure about the Kittiwake but I have been recently looking at ETW's Kiwi Mk2 drawings and ME articles. He seems to have gone to great lengths on that engine to incorporate bolted-on bronze counterweights on the steel crankshaft, presumably because of their greater mass per volume. So I would expect that his dimensions for the bronze counterweights would be correct size for correct balance. But, no idea if the Kittiwake is the same?

[Martin Hamilton 1Participant @martinhamilton1]

I always used 55% balance ( 50% – 60% is good ) on my 2.5cc & 3.5cc racing 2 stroke glow engines that i made & used to race at around 30000 rpm. Weigh top of rod, then weigh bottom of rod. weigh complete piston assembly ie piston, rings if used,wrist pin,.circlips. Then add the weight of the top half of the rod to the weight of the piston assembly, now take a percentage of 50% – 60% of this weight & then add this weight to the bottom half of the rod. This is the weight that you need to machine a simple round balance weight that will slide onto the crank pin, then sit the crank on 2 parallel knife edges to check the balance & take appropriate material from the crank web to balance. On my particular engines i also used to have to add tungsten slugs opposite the crank pin to get the accurate balance i was looking for at very high rpm. You wont need this on much slower rpm engines though.

[HopperParticipant @hopper]

30,000rpm!! Crikey. It must work then.

The old Kittiwake might not quite make that, even though it was a racing engine in its day.

[SillyOldDufferModerator @sillyoldduffer]

Posted by Roderick Jenkins on 02/05/2020 10:02:18:

… A simple single like the Kittiwake cannot be perfectly balanced but 50% of the reciprocating weight seems to be the consensus. Since Westbury designed the Kittiwake to be a performance engine for a hydroplane I suspect he probably got it balanced as best he could.

Stay well,

Rod

On the plus side Westbury was a professional engineer and I think it's safe to assume he knew what he was doing. He would have calculated the static balance, which will correct a good chunk of the balance problem. More difficult is optimising the dynamic balance because it changes as the engine rotates. Working nearly a hundred years ago, Westbury would have done a simple dynamic analysis (he had access to professional data), and then tweaked the design after measuring engine vibration in a prototype. Almost certainly this approach means the Kittiwake is approximately balanced at one particular operating point, probably not ideal over the engine's entire operating range.

Nowadays being able to number crunch dynamic sums on a grand scale with a computer makes it possible to do a much fuller analysis. Dynamic balance can be determined in many different conditions, and all the parameters can tweaked to find optimum answers. Not done by eyeballing, guessing, or a few sums on the back of envelope.

Might well be possible to improve on Westbury's design, but whether we have a Model Engineer capable of doing the work is moot. We have people capable of designing engines from scratch, but I can't think of anyone applying modern technique to engine design. This would include using FEM to rationalise strength, calculating dynamic balance, thermal effects, and applying CAM technique to minimise cost. Very likely the design would be multi-disciplinary, involving materials compatibility, simulation, electronics and a computer controlled Engine Management Unit. It's a lot of work, and I suspect the result wouldn't be a massive improvement over Westbury.

Must admit to being a Westbury fan. I find he writes at least as well as LBSC and seems not to have developed fixed ideas and ego issues as he aged. While LBSC's writings range from excellent to poor; Westbury seems to have been consistently good throughout. But of course, they complemented each other because Westbury wasn't interested in steam!

Dave

[Former MemberParticipant @formermember32069]

[This posting has been removed]

[John ViettiParticipant @johnvietti39406]

Thanks for all the input! Obviously I was only doing it half right. I think I’ll try 55% of the reciprocating weight plus the rotating weight. Finishing up the piston now. Will be interesting to see how close it comes to Westbury’s design.

John

[Howard LewisParticipant @howardlewis46836]

John,

Do be aware that by balancing the Crankshaft, con rod and piston, on a single cylinder engine.you may well achieve good balance in the vertical plane, but you will then have the secondary out of balance forces, acting in the horizontal plane.

A vertical engine,with two or four cylinders will be better, but it still tend to squirm, because the forces acting alternate from one end of the engine to the other. Incorporating a Lanchester harmonic balancer improves matters, but introduces the debate about where to site it, front, centre or rear. to minimise any residual tendency to squirm, (because the forces although they may be equal, do not coincide in their plane! )

An engine with the cylinders set it 90 degree Vee form would be better than a horizontally opposed engine.

Hopefully, the primaries from one cylinder will be matched by the secondaries from an adjacent cylinder.

(As used in the Brough Superior motor cycle).

I am not saying "don't balance" but just be aware!

Presumably the debate about what percentage to balance is to try to produce a compromise between vertical and horizontal forces for an acceptable level of vibration.

Howard

Edited By Howard Lewis on 06/05/2020 12:23:15

[duncan webster 1Participant @duncanwebster1]

90 degree Vee twin with balance weights for 100% of one conrod/piston has good primary balance, but the secondaries are not cancelled out. 270 degree parallel twin with 2 balance shafts can achieve good primary and secondary balance.Harley's at 45 degrees and Vincent's at 50 degrees are a compromise to fit the available space. Ducati and Moto Guzzi got it right

No advanced maths required, a bit tedious to calculate the various masses if you don't have a 3D Cad program, but just geometry. In reality you'd have a good idea of the weight from similar engines, so make a prototype and weigh the bits before finishing the crankwebs

[Tim StevensParticipant @timstevens64731]

Back in the days when Velocette made motor bicycles, I asked the local dealer about balance. He went to his desk and pulled out a cloth bag sewn up with some nuts and bolts inside. 'Set the crank up on knife edge bearings', he said 'and hang this in the little end. That will save you hours of experiment. This one is for the 500, I have a 350 one too'.

Those really were the days, when you could find a dealer who really knew what he was talking about.

PS Guzzi and Ducati got it nearly right – there remained a rocking couple because the big ends sat side by side. And while the Guzzi was across the frame, the inline Ducati engine went round backwards and so counteracted some of the torque reaction …

Cheers, Tim

[John OlsenParticipant @johnolsen79199]

This is a topic that can easily get very deep, and I can see a few misconceptions have crept in already.

For a single cylinder engine, there is no way to get perfect balance, unless we start adding extra balance shafts. The rotating masses can be fully balanced, but the reciprocating masses can only be balanced by a similar reciprocating mass, or by adding a pair of counter rotating balance shafts. The usual approach is to fully balance the rotating masses and partly balance the reciprocating masses. This reduces the up and down forces at the expense of adding forces at right angles in the plane of rotation. (Fore and aft on a motorcycle for instance.) The balance factor that works best will depend on the details of the rest of the structure, for instance on a motorcycle fore and aft may be less annoying than up and down.

The next problem is if the balance weight cannot manage to be in the same plane as the unbalanced forces, for instance on an overhung crank. This gives rise to a couple, even if the forces were otherwise equal and opposite. A

typical example would be on a 180 degree twin, eg most two stroke parallel twins and some of the Honda twins. (The CB350 had a 180 crank while the CB175 had a 360 degree crank.) The primary forces are equal and opposite to those from the other cylinder, but since they are out of line with each other, they give rise to a rocking couple, which is not ideal in a motorcycle frame. This is often referred to as dynamic balance, since it would appear balanced on a static knife edge testy but will vibrate like mad in service.

By adding more cylinders we can get rid of the rocking couple, for instance a four cylinder in line will cancel all of the primary unbalanced forces, as will a 90 degree V twin. This leaves us with the secondary forces. These are the result of connecting rod angularity. Because the connecting rods are not infinitely long, the pistons do not move with a simple harmonic motion. If you plot the waveform of the motion, it is more pointed at the top dead centre than at the bottom, and the force is slightly higher there. A Fourier analysis would show that the waveform contains even harmonics, eg at twice, four times, and so on of the original frequency. These forces cannot be balanced out by any weights rotating at the original crankshaft speed. They can be eliminated by cunning design. For instance, a horizontally opposed twin would eliminate them because the secondary forces from each cylinder would be opposed. Unfortunately, the horizontally opposed twin has a problem since it is not practical generally to have the cylinders on exactly the same axis, so that both the primary and the secondary forces are out of line and so create a couple.

Another consideration with all of this is trying to keep an even firing order. Uneven firing orders lead to torsional vibrations in the drive train, so even if you balance was otherwise perfect the result might not be as smooth as desired.

Brough came up with a way around this. You have two crankshafts so it looks like two horizontally opposed twins one above the other. However the connecting rods on each crankshaft share a common pin, so they are not opposed in the normal sense. The two crankshafts are geared together, so they turn on opposite directions, and all four pistons end up moving together, eg the left pair are at top dead centre when the right pair are at bottom dead centre. You would think this would give rise to massive imbalance, however each crankshaft has balance weights to fully balance both the rotating and reciprocating forces. These would then give rise to forces at right angles at mid stroke, but since the crankshafts are rotating in opposite directions, those forces balance each other out. The secondary forces also neatly cancel each other, and the firing order is even. There are other arrangements that can work, especially of you allow more cylinders.

John

[HopperParticipant @hopper]

Did Brough ever make a 90 degree twin? All the JAP V-twin engines were 50 degrees. So those would have been old boneshakers in reality, calmed down only by the low compression and low rpm of the era.

Vincent designer Phil Irving told Brit bike manufacturers back in the 50s or 60s they could smooth their vertical twins out by simply changing from a 360 deg crank to a 270. But they didnt believe him — until the modern Hinkley Triumph 270 twins. Doh.

[John OlsenParticipant @johnolsen79199]

So far as I can recall, the Brough V twins were bought in, rather than making their own. I could be wrong, and I don't recall which suppliers they used, but there were several suppliers back in the day, including Matchless and JAP for a start.

Phil Irvings book on motorcycle design is a good read, a bit dated now of course. His opinion on Featherbed frames is likely to upset a few!

John

[John ViettiParticipant @johnvietti39406]

Thanks for all the input. Used the methods suggested. Turns out Westbury's weights as designed weighed 67 grams each. After balancing exercise the resultant indicated weight is 38.7 grams, big difference. I will use this weight but worried since its so different from what the designer suggested. Won't know till I try to run it. Currently working on magneto mount and drive, starter drive complete.

Will report results.

John

[Neil LickfoldParticipant @neillickfold44316]

The important part is that the crank will sit a 12 o clock in line with the cylinder when vertical. As for the ratio that changes depending on the application and cylinder position. Some motor cross engines were balanced as low as 30 percent when the cylinder was pointing forward. This gave very little side ways or vertical vibration to the rider. F2C diesel team race engines are between 30 and 40% balance, as it makes for a lighter over all engine. It seems that the underbalanced engines are the ones with better range on the tank of fuel. The balance when from 30% to 60% aprox, when held in a very firm mount, will produce the same amount of total power on a test stand. But in a model however that number can vary hugely depending on how well supported the engine is held. In some situations it is preferable to have an engine that has more cylinder in line out of balance and less side ways induced forces.

[Oily RagParticipant @oilyrag]

John Olsen's contribution is an excellent summation of the problems of balancing engines. One aspect that has not been mentioned is when the cylinder is 'Desaxed' as this has the effect of changing the effective reciprocating masses and the change brought about is best looked at in time domain rather than angle domain. To effect a true picture of the OOB forces it is necessary to plot a loci of the rod / piston CoG point. In a non Desaxed engine this appears as roughly egg shaped, in a Desaxe engine it is a 'bent egg' shape as the TDC and BDC points are not opposite! This is where the benefit of Desaxe lies – that the downward strokes are 'longer' and the upstrokes are 'shorter'. In a 2 stroke this gives assymetrical timing of the induction, transfer and power and exhaust. When employed in a 4 stroke it extends the induction and power phases and reduces the compression and exhaust phases.

This was employed in 2 stroke engines originally by MZ and was as a result of discoveries during the war work for the Luftministrie by Walter Kaaden (looking at cheap, disposable(!) but powerful 2 strokes for the V1 flying bombs), After the war he used it in designing the MZ race engines (and was the first to use 'acoustic' expansion chambers with stinger tail pipes). The knowledge was taken by the MZ East German rider Ernst Degner, when he defected to the Suzuki factory in Japan.

The other important requirement is knowing the mounting method of the engine into the vehicle as to what is the best way to balance the engine. Sometimes more fore/aft OOB is preferable in a single if the mounts are at the base of the engine and/or just above the crank centreline, for engines mounted by suspending in a frame from the top of the cylinders the preferred OOB desirable is to be evenly split between fore/aft and vertical.

An excellent example of a really well balanced engine is the Ariel Square Four. 180 degree cranks that contra rotate. Also used in the Suzuki Gamma RG500 – main benefit in the Gamma was that it allowed for disc valve induction as an added bonus to a silky smooth engine.

Edited By Oily Rag on 22/06/2020 21:18:15

[Author]

Posts