Drivetrain design

[MATRACER]

I searched the forum but didn't found exactly what I wanted to know so here I go!

1- to select the proper chain sprocket dimensions, what should be use?

For ecemple, in the EK catalog (http://www.ekchain.com/images/EKcatalog.pdf) if I take a look at the 428SROZ, it is written that I should use this for a 15hp application and that it can wistand 5300 lbs. Our max chain tension will be around 2200lbs but we will have a lot more that 15hp... Should I only consider the tensile strengh because the car is not designed for 10000 miles?

2- what is the speed that should be used for selecting the final drive?

I saw number for 2004 and 2005 event, since michigan state did high three in the accelaration and that the track at fontana was fast (I don't know about Detroit), I guest we should design for faster top speed, but how much ?

THX!!

Mat from Sherbrooke

[Composites Guy]

1.a. You could try to estimate the load in the chain during accelleration based on the max torque for your engine. Look up a dyno curve for the bike you took the engine from, and be sure to account for the different tire and sprocket diameters between your car and the bike.
1.b. You could also estimate the load in the chain as the tire slips at the begining of acceration. Estimate car weight, wheelbase and vertical CG height, and get the weight over the rear tires. Take a conservative (high side) acceleration of 1.5 g. Use tire diameter and rear sprocket diameter to get the load in the chain.
1.c. Realistically I'm guessing that the estimates above are too low. I'd think that the actual highest chain loads happen when the clutch is dropped, or non-smooth shifting happens. This slamming load would have to act not only on the friction capability of the tire but also the inertia of the wheel, axle, diff etc. For this reason I'd give yourself a healthy factor of safety over the estimates above.
1.d. Look at what other teams are using. Mainly 520 chains. Some teams run 428. We have tried this in the past but found the chain stretch and sprocket wear were too sketchy.
1.e. Search the forums to see what other teams are running for front sprocket # of teeth. Run the mininum you can get away with without excessive wear. Choose what top speed you'd like to see in your gears, and let that determine your rear sprocket diameter. Making a quick spreadsheet helps.

2. That's a matter of choice. The top track speeds at detroit are usually ~65mph.

[MATRACER]

We will be running a CVT. I already calculated the maximum chain load with a certain sprocket and it will be around 2200-2500lbs. That would give us a safety factor around 2 for a 428 chain. It used the maximum estimated torque of the engine. Doing that, I don't consider the clutching impact but I maybe overestimate the real transmitable torque.

There will also be some efficiency lost ine the CVT, so the really maximum deliverable torque is probably under what we estimated.

Because we are running a CVT that is best suited to run at high rev, the final ratio must be around 6:1 (155Kph theoric max speed) or higher. With a 520 chain, we would get a pretty big diff sprocket!


Has any team encounter a sudden break with a 428 chain? Wear is not a real issue since the car is design to run a 100 miles... http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif

THX!

[rjwoods77]

I have seen some teams use 428 on 4 cylinder cars and survive so that might be worth something to consider. The whole hp capacity of chains never seems to match up even when you are using ANSI chain. I think it has more to due with industrial guidelines much like b10 life for bearings as I have seen chains that are only supposed to support 5hp hold 30hp.

Just for something to think about your baja team did something in 2001 that was a very good solution to your problem. They wanted to use a single gear reduction instead of the normal two stage to reduce losses,costs,etc. Baja gear reductions are typically anywhere from 7:1 all the way up to 12:1. If memory serves me correct Sherbrook had a 8:1 reduction they did in one stage. If they used 428 or 520 chain the sprocket would be huge and wouldnt fit (approx 12") so they instead used a tripple row ansi #35 chain so they could get a smaller diameter final drive sprocket due to the smaller pitch distance while at the same time having enough chain strength to comp with the crazy shock loads you get in baja. Something to consider in packaging. I hate ansi chain myself because of the crazy stretch you get in it but there are definite benefits to using it since you can get it anywhere. Especailly connecting links. Sherbrook baja rocks!!!

Also something to consider is a cvt will absorb a certain amount of shock as long as the chain is stronger that the shock load and that tension is greater that the clamping strength of the sheaves. Its like a chain strain tensioner if you understand that. Same as if a clutch on a shifter car were to slip under a heavy load.

[Chris Allbee]

Unless you are using a torque converter, I wouldn't ignore the impact loads of the clutch drop. And for that matter the "jerky" start of less experienced drivers. These impact loading are ridiculously large (2-3 times in some cases)compared to transmitted loads during normal operation. This can lead to excessive wear and premature failure. Also, be careful about designing the car to last just through competition and then tacking on a 1.5 factor of safety. With such thin margins Murphy will surely find a way...

[MATRACER]

Rob, I agrre with the Skerbrooke's Baja ROCK! I have talked a lot with the baja's guys and they are still using no35 triple chain with a 9/70 ratio.

Look at this catalog:

http://www.ustsubaki.com/pdf/gen_catalog/gen_a6-a12.pdf

the RS35-3 chain has an minimum ultimate Strength of 5280 witch is approximatly the same than a 428 chain. But industrial chain, in my mind, are not design to wistand the high speed of a bike chain. The baja is going 60 kph max. I will look at the max speed that that kind of chain can take.

A single stage gear reducer would force us to do a 180 degree with our engine and it would really give us a bad packaging.

Has anybody ever tough using a belt drive reducer like the one used on Harley?

THX

[rjwoods77]

Formula cars on the track are only going about 30 percent faster than a baja car so I wouldnt worry about that too much. UB baja car has been hitting 45mph. The problem with belt drives is that there is a very limited selection of big tooth pulleys and the pulleys themselves are much more expensive than a sprocket.

[SR-Mike]

Although you may want to design your car to have a life no longer than what is required for one years testing then competition, i would highly recommend that service requirements are not ignored entirely. I would therefore suggest that on these wear components such as chains and sprockets that you integrate as many of the shelf motorcycle (or similar) components into your designs as possible.

Advantages Include:

1 Many of these components are designed and manufactured for similar operating loads and conditions as are seen in most FSAE cars.

2 If replacement parts are required then they can be purchased cheaply and delivered within a matter of days.

3 For the rear drive sprocket this also opens up the possibility to not having to fix your final gear ratios during the design phase, but allows you to buy a number of components that allow you to then test a range of final ratios and chose the most suitable based on those tests. Remember ratio optimisation calculations during the design phase rely on big assumptions such as the track layout and conditions, both of which are very difficult to second guess.

Saying that you might struggle to get up to a 6:1 ratio with stock parts.

[Composites Guy]

I agree with Chris Allbee. The impact loading is going to dominate. I myself look at the static loadings I mentioned in my post above, pick the largest static loading and multiply by a factor of three to get my design load. I have no data to back up this approach, except to say that this results in parts that have worked without braking or yielding in the past.

Also... if you use FEA, work out the deflections due to the loading. Figure out if these deflections will cause significant problems (misalignment of the chain for instance.)

[JD232]

I dont know guys, why you all have such a supreme trust in a chain drive. I personally find them very treacherous.

I am working on a transaxle design, I dont have enough data yet to say anything for sure, but no more chains to be sure.

[Big Bird]

Well I have just spent the last weekend down at Phillip Island watching "treacherous" chain drives flinging MotoGP bikes around at 330kmh. I'd think you'd need a bit of trust in your equipment to do that.

But if you want to spend your time effort and money designing a one-off transaxle, rather than lay down $100 at your nearest bike shop for off the shelf components that will do the job perfectly well, then go nuts. I just think there are much more effective ways to spend your time.

Cheers,

[Underthefloor]

How much weight will you save with 428 chain over 520?

[rjwoods77]

Non O-Ring:

428 2.2lbs/100 links
520 3.54lbs/100 links

[SR-Mike]

Remember the pitch of a 520 and a 428 are different, so if you want to compare the mass you are better comparing lengths and not the number of links!

100 "428" links = 100*1/2inch = 50 inches
100 "520" links = 100*5/8inch = 62.5 inches

Also, because a 428 chain is wider, the rim of both sprockets should be thicker, = more mass.

Mike

[HenningO]

Originally posted by SR-Mike:
Also, because a 428 chain is wider, the rim of both sprockets should be thicker, = more mass.


But with a 428 chain, the diameter of the sprockets are smaller(with the same # of teeths) = less mass

[Brett Neale]

Originally posted by HenningO:
But with a 428 chain, the diameter of the sprockets are smaller(with the same # of teeths) = less mass

But when your front sprocket is already as small as physically possible for the output shaft, same diameter = more mass...

[A Richards]

You dont save weight running less teeth, you still need the same ratio of the PCD front to rear to maintain desired gear ratios. We make our own sprockets and on the front we are able to run 12 teeth at 1/2" pitch around a 25mm output shalft. 49.08mm PCD is the smallest we can run. The size of the rear sprocket will not change for the same gear ratio. We only have 6mm thick sprockets for 428 and have not had any problems.

[Underthefloor]

We compared the actual weight of a 428 and a 520 chain a few years back. I don't remember the specifics but the two chains were much closer in weight than we expected.

Can anyone weigh two of these chains and post the numbers? It would be interesting to see how accurate the spec sheets are.

[VFR750R]

Smaller sprockets (higher chain tension) with a lower strength chain sounds like a bad bet to me.