Rod to Stroke Ratio

[Neal H]

Hi Guy,

I was wondering if you can help with an engine selection issue, well, more specifically an engine block selection issue. A little background: with 80's and 90's VW 16 valve engines two different configurations were produced, both with a 92.8mm stroke. One had 144mm con rods, the other 159mm rods. This gives us the following rod to stroke ratios: 1.55 and 1.71

My question is, is there an ideal ratio that should be used, and what impact, if any does this ratio have on power and torque delivery?

Thanks for any advice in advance,

Neal.

[dflinkmann]

I would be also interested in the experiences in a general focus (so not only Fiat and VW) on this.

Daniel

[Guy Croft]

The longer the rod the less early the advance needs to be, so the piston is not fighting the expansion of the burning mixture for so long. B

By that I mean ordinarily the ignition might typically need to commence, depending on the efficiency of the combustion cycle of a given engine (dependent on piston crown design, combustion chamber design and number/placement of valves) anything form 28 deg to 35 deg before top-centre. With a longer rod the piston 'dwells' around TDC for longer and thus the ignition can start several degrees later and yet still give complete burn in time to develop maximum thrust. This can, I do know thru incontravertible (reliable) third party feedback, a measurable increase in power.

How long you can make the rod depends in part on aval piston forgings with short height from pin to crown and of course lengthening the rod (given that it's steel) can often lead to an increase in mass of the upper rod half/piston assembly, so it is just possible that the increased inertia of that region could undermine the a gain in power from the late advance. The other thing to remember is that the longer the rod the higher the bending stress - an issue for high boost units. Just because they do it on some race engines doesn't mean it's right for you. Guys who take a chance on very long rods may well junk the rods every race, I don't know.

Remember too that the longe rod has to be checked in the crankcase - it may foul.

Hope that helps some.

GC

[Neal H]

Fantastic, thanks for that Guy. Do you know if there is an optimum ratio to aim for, or just a generally accepted range to be in between?

Thanks again,

Neal.

[Guy Croft]

Neal hi

There isn't an optimum that I know of, no. This stuff is probably very much trial and error though advanced engine simulation can predict the behaviour without the need to run the engine - remembering always that:


1. The minimum length of the rod will generally be that needed to put the piston in the right place in the bore at tdc and clear the crankcase.
and:
2. The length of the bore is a function of engine cubic capacity.

GC

postscript:
The 2 liter TC Fiat has 90/145 which is 62% and I would like to try 90/155 or even 160 which is 60%, 56%. Struggles to get over 100bhp per liter.
The 1438 TC (which is a dandy little motor) has 71.5/128.5 which is 56%. Flies past 100bhp per liter.


Rod length is not generally published so if anyone would like to add to this list we may see a picture emerging. I'd be interested if anyone (reliably) knows the stroke/rod length and bhp on any Touring Car/F1 engines.

[1NRO]

As Guy mentions the increase in dwell is a handy symptom of a longer rod. A longer rod reduces peak piston speeds slightly. The peak velocity is delayed until the piston is further down the bore, gives the intake valve a little longer to open before peak airflow. Reducing these peak velocities assists in the tensile loading on the rod. A longer rod also reduces the maximum rod angle which in turn reduces the side loading on the piston which helps friction a little, this is linked also to the reduced pin height for the piston which reduces the piston rock slightly. All tiny improvements I admit with things like a reduced ring pack to worry about. There are arguments for and against using longer rods, it isn't all rosey. I personally think it is a worthy route to follow with ambitious engines.

145/90 = 1.61:1
160/90 = 1.77:1

A target of 1.80 is considered handy in high rpm applications with the likes of F1 much higher still. Can't quote what they do use as this isn't freely shared, suprise suprise. Over 2.0:1 and possibly as high as 2.2:1 I'm told.

[Guy Croft]

Some good points Nicholas,

I've examined the effect on velocity with long_short rod and it's absolutely miniscule. I had to lengthen rods by huge amounts to see an effective difference. So in terms of effect on valve event it can definitely be ignored. See att.

Definitely? I can't believe I said that..!

GC

[1NRO]

It is miniscule, I agree. There's lots of tiny variables when changing rod length. I like the improvments in the piston area mostly. The low rpm range doesn't bother me so much as the high end. A move in the opposite direction has benifits for lower rpm, whatever suits the purpose. A production specification is production compromise with consideration for factors that might not concern the racer.

I wonder too about the stroke, that isn't as easy to play with though.

[1NRO]

The graph is interesting, thanks for posting that. I've never seen it like that. The area after 90* and before 270* particularly. Is that calculated at a particular rpm?

[Guy Croft]

8000 rpm in both cases, halve the speed and you halve the piston velocity but they will still develop peak velocity at their same respective crank angles.

GC

[Neal H]

Looking at the graph above, piston velocities are higher with shorter rods, pre 90 degree crank rotation, and post 180 degrees. Would this not result in lower pressure in the cylinder earlier in the stroke, resulting in a higher VE (raising the kinetic energy of the charge earlier in the valve event), or am I reading this wrong and talking total poor?

[Testament]

On this topic, Guy have you ever heard of anyone running a 80mm 1756cc crank in a 1995cc block? it would be possible to fit 160mm rods (2:1 ratio) without squeezing the pin up too high in the piston. I am guessing an 8v TC head wouldn't really be able to make use of the rpm capability and 16v would be a necessity.
Any comments on this possible setup?

[Guy Croft]

Regarding the influence of rod length, piston vel and VE - this is not something I have had time and funding to explore in any detail, so I cannot regard myseld as expert, but I am certainly of the view that a key parameter is getting inlet FL as close to peak piston vel as possible - of course within the constraints of the piston-valve relative motion. Obviously we don't want tehm to hit each other around TDC! That kind of makes sense to me.
Unfortunately there are so many facets to cylinder filling on a reciprocating unit it's hard to know which is more important and over and over I find myself just having to fall back on good ol' experience.

For example, we know that the inlet charge comes in because of a pressure differential between the cylinder and the port. The higher this Rp the faster the charge will enter (after the inital choking phase at the valve is behind us). So, why not delay the valve opening until the cylinder depression is at a maximum? Because doing that would spoil the overlapping at high rpm that engines need - where both valves are open - and the exhaust depression encourages the incoming charge. At least that is our present best understanding.
Moreover if you hold the valve even till very late in the inlet stroke you are vastly reducing the time available to fill - remember air/fuel charge has inertia and, given the choking at low lift etc you can't expect it to respond instantaneously. Then there is the strain on the valve train, the accelerative rate would be enormous, more than a spring (or similar system) could cope with.

Those are just some of the things you have to think about and naturally better brains than me have sought, over generations to turn a powerplant that is essentially no more than a steam engine running on gasoline into something superb - and failed.

I had a particular Virtual 4T simulation done where the software suggested 95 deg inlet FL (2.3 HS 16v Vauxhall). That was quite a shock and I didn't have the nerve to build it that way - the lift at TDC would have been enormous, 7mm or more. Could I have reduced the LATDC? Well maybe but then the software would almost certainly have suggested FL in a different place, so I just settled for a more contemporary 100 deg ATDC, 5.2mm.

Insofar as putting an 1800 crank in a 2 liter, no I have never done that! But in terms of our present discussion it might work, might it not? If the rods didn't foul the crankcase so much that you broke thru trying to get them to clear. One would have to assume the crank was actually strong enough, and you'd certainly want the Fiat 132 1800 crank with the 12mm FW bolts, not the 124 one with 10mm...

GC

[Neal H]

Thanks Guy. Taking this totally theoretical discussion one step further, for a short rod setup, peak valve lift probably needs to be earlier to coincide with peak piston velocity. The point at which peak valve lift occurs is determined by a number of variables including cam design and the ability of the valvetrain to cope with the desired lift profile. For arguments sake, if a cylinder head has a peak flow at a very low lift, a cam could be designed within the capability of the valvetrain to open early enough to take full advantage of the short rod engine's earlier peak piston velocity, and the opposite for a cylinder head that requires lots of lift?

[Guy Croft]

That's true but only in the most general terms, because it is in the nature of a poppet valve to restrict the flow severely at lower lifts, that's the discharge coefficient at work against us, and typically it's necessary to lift the valve anything from 25% to 40% of valve diameter (on an optimised head) to take advantage of the maximum flow potential of the valve. Or put another way you have to lift poppet valves really quite a lot to achieve a position where the valve head no longer presents an intrusion into the flow.

Of course it's quite true that the flow either_side_of_peak is important but the massflow at peak lift is hugely important because of the high rate of fill at that phase in the valve event (and I have proved this over and over with pretty demanding flowbench dev and back-to-back dyno results) and the most powerful cams have significant dwell at full lift to capitalise on that.

Apart from that every unit has to be studied on a case-by-case basis.

G