Intermediate Shaft Straight-Cut Gears, real deal?

[Jimmy18E]

Guys,

Is there any real advantage to Intermediate Shaft Straight-Cut Gears?

I have read everything I could find with the "search tab" and no difinitive answer.

I am collecting parts now for the build and was debating if the $1100 is worth it or go with the stock gear? It is in a pretty maxed out motor for fast street/DE, noise is no issue.

Thanks,

Jimmy

[304065]

They are stronger than the helical gears.

They will also make a noise like a jet engine. The only application is pure competition where you are wearing a radio harness with earplugs and a helmet.

[ajwans]

Originally Posted by 304065

They are stronger than the helical gears.

They will also make a noise like a jet engine. The only application is pure competition where you are wearing a radio harness with earplugs and a helmet.

Is this true? I read that helical cut gears are harder to break (effectively stronger)
because they have more than 1 tooth engaged at any one time. Isn't the real
advantage of straight cut gears that you transmit more power perpendicular to
the shaft because there is no lateral thrust being generated by the cut of the gears?

There's a pretty good discussion Intermediate Shaft Straight-Cut Gears

andy

[jsveb]

From what I know straight gears are stronger, but makes more noise.

[Walt Fricke]

I went to straight cut gears in part because they were all steel. Stripping the teeth off of an aluminum IS gear is something one does not want to repeat. If you aren't spinning the motor up to 7-8,000 rpm all the time, probably much less of an issue. And nowadays you could get a steel helical gear (from a 964?) to deal with the strength issue.

I also operated on the monkey see principle: Porsche did this with its factory race cars in the '70s. In addition to having little or no lateral loading, it was my understanding that there is slightly less friction involved with the straight cuts. But my understanding there could easily be wrong.

On a race motor you will never notice the noise. Don't know about street.

[Walt Fricke]

Oh - I got a deal (this was 10 or 15 years ago) on these gears. At $1,100 per set I think I'd pass, and spend the $ on some other gofast.

[chris_seven]

Originally Posted by ajwans

Is this true? I read that helical cut gears are harder to break (effectively stronger)
because they have more than 1 tooth engaged at any one time. Isn't the real
advantage of straight cut gears that you transmit more power perpendicular to
the shaft because there is no lateral thrust being generated by the cut of the gears?
andy

A helical gear has a stronger root at the base of the tooth because there is a longer line length. The means that the bending fatigue stresses will be lower for a given torque transmission and the tooth should be more resiatant to fatigue failure.

The number of teeth in engagement at any time is a funtion of the Contact Ratio and varies with the number of teeth on either gear pair.

A very simple gear can have a contact ratio of 1 but this would produce very high levels of noise and could wear quite quickly of the pitch accuracy of the gears was not extremely good. I don't think this type of design is very common except at very low speeds and loads.

A more typical contact ratio would be about 1.6 where ther are two teeth in engagement for the majority of the time and this type of gear set would be quieter and more durable.

The helical gear does feed in load to each tooth in a more gradual manner that a spur gear and so the noise generated by 'landing forces' is always lower. This will also reduce the tednacy for a gear to scuff and genrate pits which will also improve fatigue life.

The side thrust generated by a helical gear is a factor to consider and this can become an issue when there are torsional vibrations being generated - as is the case in a cam drive. Having said that the aluminium gear seems reliable but with high lift race cams the torional vibration and associated fatigue loading generated by the cam could be more of a problem.

As I said in an earlier thread I am not sure that a spur gear will provide much of a power gain but they are cheaper to cut than helical and certainly more reliable.

The 964 steel gear may be the best of both.

[HawgRyder]

Some machinery use a double helical gear...with the teeth in a "V".
In transfers the power more efficiently and is fairly quiet, and the added bonus of having no side thrust.
As to power consumed...good question.
Bob

[chris_seven]

Double helicals are now quite rare due to cost and were initally produced by Citroen - hence the shape of the logo.

The will create slightly more power loss than a spur but no too much.

Having been involved in the design and manufacture of high speed rotating machinery test systems and having used a significant number of industrial gearboxes I do have some knowleges with regard to power loss.

A typical estimate for a single stage parallel shaft gearbox would be 0.5% of the transmitted power and this is generally backed up by heat to oil measurements.

Two-Stage gearboxes would be about 1% and a box with a single input and two outputs would be about the same.

I would size an industrial gearbox oil cooler by assuming 2% to allow a bit of headroom and I would differentiate between spur and helical as difference are minimal.

Bevel boxes asre a different issue and I woul estimate about a 2% loss in a bevel box and maybe 2.5% for a spiral bevel. bevel boxes are used commonly in four-square transmission test rigs.

These figures are not perfect but a decent rule of thumb for sizing drives, coolers etc,

[304065]

Chris, what a great fact about Citroen! No doubt driven by highly optimistic expectations of the development potential of the 9hp two-banger in the deux chevaux vapeur!

(Of course the company dates to before the 2CV but it is easier to poke fun at their iconic, well-engineered triumph of automotive minimalism!)

[Walt Fricke]

Chris

One of the mysteries of engine work is the factor one should use to compare chassis dyno figures to those from a flywheel dyno. The factors one sees range from maybe 10% to 15%. Most of that is supposedly in the gearbox, though some has to go to the heat the chassis dyno run is putting into the tires?

What would you estimate the loss in a Porsche 915 box to be?

I suppose this is complicated by the fact that half of the non-engaged gears are still spinning (and having inertial effects on a Dynojet type inertial dyno) and thus creating some amount of drag.

[chris_seven]

Walt,

Measuring transmission losses on a gearbox is, in test rig terms, quite difficult.

If we take a simple example of a gear box with a power transmission capacity of 100 HP and a loss of 0.5% when we load the box we will need a torquemeter of the relevant capacity - depending on speed. If we buy a high quality instrument with an accuracy of 0.1% we can only estimate power loss to 20% - not too good for the cost involved and this is why there isn't too much good data.

If we think about a 300HP (225kW) motor then a loss of 15% would be almost 35kW.

I have to say that I think this is too much and looking at transmission temperatures would give a better idea. As we know the weight of the box and can estimate the heat loss from steam tables (usually published in Engineering Handbooks) then the Delta T would give a reasonable first order guide.

The gears that are idling and not transmitting power would give some loss but this would be minimal and difficult to predict but even if it doubles the typical losses of the driving gear and the final drive we would still only get to around 5% and I think this is realistic.

I also think if a 915 did have to try to reject 35kW it would melt after a few hours

Being cynical I have tended to think that as the Rolling Road can only sell horsepower improvement that bigger the factor the better abnd here in UK figures of 20-25% are often quoted !!

Also a typical chassis dyno has all sorts of measuring errors which need considering.

Interia dynos are affected by gear ratios particulalry on cars without a 1:1 gear ratio in the transmission, wheel slip and tyre losses are also an issue.

The poor torque measurement systems often used are also another source on uncertainty as is the general lack of calibration.

The last chassis dyno I looked (not a Dyno Jet) at in detail had an electromagnetic truck retarder mounted in a trunnion as the load. The torque measurement was provided by a simple beam with a couple of strain gauges. The beam had an L to D ratio of about 1.5 which is nowhere near good enough for a measuring transducer and must be very non-linear. The losses in the trunnion bearings are also significant. A 1000HP chassis dyno would commonly give error of 1% of range so the lower the power measurement the worse it gets.

I am not sure how the Dynojet measures torque but I have seen some machines that use current measurement and take no account of phase angle and other sources of error.

In general my conclusion has always been that it is almost impossible to compare engine dyno and chassis dyno results with much confidence and it os also difficutl to compare on engine dyno with another - assuming you are using either water or eddy current brakes.

GE, for example, have a 'Golden Engine' - typically a T-700 that they use to check a number of test cells used forTurboshaft engine evaluation so they gain confidence that they all produce the same data.


The typical modern F1 engine dyno (commonly supplied by AVL) uses AC motors to brake and drive engines hence simulating inertia during gear change and also have accurate and in line torque measurement and these systems are both repeatable and comparable from one to another.


For really serious work I think I would test an engine on a good 'engine' dyno - measuring temperature and pressure and then use a given chassis dyno to obtain a baseline and then use the baseline to look for changes rather than consider the data as absolute.

We recently bought a 3 axis acceleometer to mount in different cars and when used carefully can give quite a decent estimate of power without some of the risks associated with chassis dynos and is very cost effective.

Sorry to bang on but an interesting set of problems that have no easy solution.

[PFM]

Chris,

Nice write up and I agree the numbers from engine dyno to chassis dyno or even chassis acceleration rate dynos seldom share any real common numbers. When two or three same make same model engine dyno's do not match go figure.

Even if you manage to attempt to factor out humidity, temperature and baro the dynos seldom line up. As you said in your post these are tools to tell you if a change made was for the good or not (trends), hopefully in a back to back same day same dyno tests. However if you move from one dyno or dyno type to the next and expect to follow a trend be prepared for a tough time.

The other dyno folly I see are making multiple changes or adjustments between pulls, so a small jet change and a change of two degrees of timing, the jet was wrong the timing was right or vice a versa and the two gave a net zero change or a slight loss in power. Had they been done one at a time the true results would be reveled. The other fun is "knowing" what the air fuel ratio needs to be for best power and tuning for this number (EFI guys), truth is it may be close but seldom is "the number".

[Walt Fricke]

Chris

You are better placed to understand dynos than I. But I believe the Dynojet is not a brake dyno of any sort, so you can't hold the throttle at some degree of opening, manipulate the dyno's brake to keep at a certain RPM, and go from there to make calculations.

I think it simply depends on the time needed to spin up the drum's very considerable mass, calculating for each so many degrees of increase in angular velocity as it goes along. One dyno guy I have worked with won't let me do the sitting in the car part, lest I mistakenly put a foot on the brake! He thinks that might break his restraints and put me into what he has stored in front of the dyno.

Which isn't to say this solves all the issues. For better or worse, Dnyojet has achieved a reputation for repeatability and ambient effects correction to the point that some amateur race organizations (NASA in the US in particular) class modified cars based on their dyno sheets.

[HawgRyder]

OK..quick question along the lines of power to the cams.
Has anyone tried a cog-belt system?
I know the offset would be dependent on the width of the belts at the crank...but aside from that..would the belt last?
edit: Ooops...just realized that you could drive both cams with one serpentine belt.
Bob

[PFM]

HawgRyder,

Belts would need to be same width and I think two because of space limitations but the belt would run in a nasty oil bath most belts are built to run in open air.

[Walt Fricke]

You would have to have two belts, because the IS shaft gears are offset. Because the cylinders are offset, and thus so is whatever gear you would put on the end of the cam.

Leaving aside the question of whether a belt would last in the oil, why would one want to do this? Save a few ounces?

[HawgRyder]

Just thinking..(product of retirement and up too late at night...LOL.
OK...how about a train of gears instead of chains?
Was looking at a spare case today...and I think with a little engineering...perhaps only one gear in between the crank and the cam...but it would have to span the distance in between those two gears.
It would eliminate the chains..the ramps..the idlers...and several other small pieces.
It could be aluminum...ball bearing hub...and straight or helical teeth.
To keep the same direction of cam travel...it would need to be an odd number of gears...so total of 3...or 5 gears.
Like I said...too much coffee...not enough to keep me busy...LOL
Bob

[Walt Fricke]

Bob - you are right in line for a job with Porsche in their engine department, circa 1955 or so. You have sort of reinvented the 4 cam Carrera motor for the 550 spider and the 904, seems to me.

These reportedly took a full day or two to time, if done by an experienced factory mechanic. Astounding motors, but I don't think it was all those gears which made them so powerful.

[chris_seven]

Originally Posted by Walt Fricke

But I believe the Dynojet is not a brake dyno of any sort

Walt, I agree that the Dynojet is an interia device and hence can't hold steady state which is a limitation and is affected by gearing but this doesn't mean it may not be repeatable for a specific set of conditions but it is also difficult to take the values as absolute as calibration of these type of devices is quite tricky.

I do think that it is important to think of all dynos as a brake of some type as the engine needs a load to operate against. In the case of a basic Dynojet the inertia provides the load and as you say the torque figure is calculated from angular acceleration multipled by a proprietray Bulgarian Constant. I would think that the accuracy of angular measurement and the time constant used for integration are also important.

The Electromagnetic Retarder that can be fitted to some Dynojets units and is standard on others is a relatively cost effective method of allowing steady state tests to be made and the unit supplied by Dynojet is said to have an integrated torque measuring system but it is hard to make much more comment as the information on their website is limited.

I am much more familiar with either Sun of Mustang Chassis Dynos which do allow steady state measurements and have full capacity retarders.