Torsion damper in the drive train

Status
Not open for further replies.

mdt-son

VIP MEMBER
Joined
Jan 19, 2012
Messages
1,934
Country flag
Gents,

Not wanting to mess with the thread on shockproof oils, here is how Ducati solved their need for a torsion damper. Their design is hardly different to the AMC design we know which was in use right up to 1968, just bigger and probably better tuned.
The beauty of using rubber is that the functions of a spring and a damper is obtained within the same volume of material.
A similar solution should fit into an enlarged Norton clutch center which would still fit into the clutch (chain- or belt-) wheel. The solution would need tailored clutch plates.

I'd like to hear your views.

Pic 1
Torsion damper in the drive train


Pic 2
Torsion damper in the drive train


-Knut
 
Last edited:
Wonder why Norton did not come up with a construction like this, instead of the marginal cush drive in the rear hub.
 
Is that not like most brit twins with the paddle and rubbers? Enfield had one PLUS one in the rear hub. Trident has one inside the gearbox chainwheel although the clutch is a totally separate piece.
I can never quite understand why the Commando has no cush at all or at least until the later ones with the three fingers. It wasn't sorted until MkIII.
 
I must admit to wondering about the need for a torsion damper. Common and others have reported doing considerable milages on damper-less Commandos, while others report problems with destroyed gearboxes etc.

I do understand the theoretical need for such a damper, and the vast majority of road bikes have one.

I have a Yamaha MX wheel sitting in the workshop which would do very nicely in the Commando, but it has no damper at all, so I'm a bit hesitant here....
 
Is that not like most brit twins with the paddle and rubbers? Enfield had one PLUS one in the rear hub. Trident has one inside the gearbox chainwheel although the clutch is a totally separate piece.
I can never quite understand why the Commando has no cush at all or at least until the later ones with the three fingers. It wasn't sorted until MkIII.

Actually it wasn't sorted with the introduction of the Mk3. It's a myth that the rear wheel cush acts like a torsion damper. The reason a say this is that you've got a long elastic spring (aka the chain) in between the transmission and the rear sprocket which reduces the tensile pulses quite a bit. Even if there was a direct coupling, the torque pulses will have to go trough the transmission (and by way of the primary drive ratio, these fluctuations are magnified by a factor of 2.19) before acting on the cush elements in the rear hub.

Torque pulses needs to be cushioned before they reach the transmission, either at the crankshaft (as HD does it on their Sportster engines) or at the clutch, as demonstrated by Ducati and others.

-Knut
 
Last edited:
My homebuild belt primary drive has a rubber cushdrive incorporated in the outer clutch drum
 
Copied from the shock proof gear oil thread:

I’m in a camp similar to ‘mdt-son’, where he suggested above that short gear life is in part due due to a quality issue with the gears. Possibly the result of the wrong alloy steel, the wrong heat treatment or both. It is certainly possible to make a high quality gear for this application that exhibits long wear life. And there are certainly gears similar to the common Norton gears made for other marques that exhibit the same sort of premature spalling gear wear we’re discussing here.

Just thinking out loud about gear hardness, perhaps a manufacturer is putting 0.020” to 0.030” of case hardening on the gear, when in fact it requires 0.050” for long term survival. The gear will always have a soft ‘under belly’ beneath the hardened surface, and perhaps if the case is not thick enough, the hard surface incrementally deflects into the softer underbelly and over the course of time small ruptures of the case occur and result in the spalling type wear pattern we see? If this is actually happening, I don’t think we can lubricate your way around such an issue.

Earlier in this thread the subject of HD and common oil in primary and tranny came up. Tying the subjects of common primary/tranny oil and high quality gears together, consider the gears Andrews makes for Harley Sportsters (read the description of the product on the right page [4 speed] at the link below, and turn the page if you want to read a little more about their products).

https://rssd.cld.bz/Andrews-Products-Catalogue/28/

Andrews simply makes the best products on state-of-the-art equipment and they perform stellar – you simply cannot buy a better gear for a HD Sportster. I’ve intentionally steered the conversation off target here (from Norton to HD) to make the case that readily available very high quality gears (that don’t spall) exist for specific marques, and feel confident that if Andrews were manufacturing Norton gearbox innards we wouldn’t be having this conversation. I hope that CNW’s gear undertaking results in a product of similar quality to Andrews, and leaves the issue of spalled gear faces behind.

In closing want to add that the HD Sportster gear box/primary drive performs just fine with ATF (Dexron II, no spalling type wear), although the factory always recommended oil, and suspect that oil would, in the final analysis, ultimately result in less wear over extremely long use cycles. The majority of teams campaigning HD XRs (same 4-speed gearbox used in Sportster since 1957, but with 90+ HP) in flat track racing also employed ATF in the gear boxes and tranny with no wear concerns (admittedly this is not a high mileage use, but is nevertheless a brutal short-term use). And by the way, for many decades Sportsters never had any form of cush-drive or compensator sprocket, yet the gears held up fine. My $0.02.

Even the initial HD Sportster (the "ironhead" engine) had a comparator (a cam/cam/spring device as fitted to british bikes of the 40's and 50's, which is a kind of torsion damper), so I guess the XR-750 had it as well. The torsion damper was refined for the EVO engine 1986. To claim transmissions of engines not having a comparator held up fine means harking back 80 or 100 years - I don't think you want to go down that road.

I am sure Andrews make fine products. However, they do rely on a working torsion damper at the crankshaft. As you say, there is no general complaint about factory gears showing signs of spalling, which indicates the HD design works well.
By the way, the V-twin has a 315/405 degree firing interval. Not that much different to a parallel twin. The big difference is the crankshaft design.

-Knut
 
No pic but I used a small jap clutch as it was laying around and accomodated the 8 rubber buffers at the rear of the clutch basket. The belt drum has 8 holes where the rubbers rest. I use a magnesium hub from KTM MX at the rear so no shock absorber anywhere and I thought it neccessary. Cannot fill pics here.
 
I agree with Knut, the damper functions better located in the clutch, rather than the rear wheel.

Slick

Better,.. Why to both of you guys?

Because you believe that the closer the elastic part of the drive train is to the gearbox, the more it absorbs shock to parts close to that point?

Do you think there's all that much difference if the elastic is in the wheel hub? (obviously you do, so I mean to ask why, and what is the difference?)

Given that the origin of the force is generated by the engine into the crank and the resistance of that force is the wheel's contact with the pavement, would there be that much difference where the elastic is located?

I could see the greatest difference being in a bike that has both an elastic clutch and an elastic rear wheel hub, but I don't think there's a significant difference where you put the elastic parts in the chain of force. I think the quality of that elastic part is more important. That's why I think the late model commando rear wheel elastics are pretty poor.

I have a yamaha rear wheel with sizable elastic rubber buffers in the hub. I think Madass's cush rubber hub is about the best elastic part made for a commando.

So,... IF you think the elastic clutch is better than a modern elastic hub, How much better is it? 1%? 5% 200%?
 
The response below has been copied from the "Shock proof gear oil" thread.

-Knut[/QUOTE]
Even the initial HD Sportster (the "ironhead" engine) had a comparator (a cam/cam/spring device as fitted to british bikes of the 40's and 50's, which is a kind of torsion damper), so I guess the XR-750 had it as well.
-Knut
Your assumption about the Harley XR having a compensator sprocket is incorrect as they never employed such a device – it was always a plain sprocket on the engine shaft (item 27 in the parts book illustration from the ’57-67 Harley parts manual). Regarding the kick-start XLCH model iron head Sportsters, from their introduction in 1958 for at least 20+ years they likewise employed no compensator sprocket. The compensator sprocket (item 26 in the illustration) was employed only on the XL and XLH models. With the passage of time and elimination of kick-start models from the Sportster offering, all Sportsters ultimately employed primary drive compensators.
 

Attachments

  • Torsion damper in the drive train
    Sportster Primary Drive Parts.jpg
    100.4 KB · Views: 334
Better,.. Why to both of you guys? So,... IF you think the elastic clutch is better than a modern elastic hub, how much better is it? 1%? 5% 200%?

See my entry #5 above.

As for improvement of one solution vs. the other, remember why are talking about torque pulses here. For a parallel twin we can expect peak values of torque at the PTO shaft of 20 % above the average value, and there will be one peak per crankshaft revolution (360 degrees). In addition there will be 1st, 2nd, 3rd degree torsion harmonics superimposing.
This "ripple" may account for another 5 % of the average value. If we neglect the dampening effect of the primary chain because of its short length, these torque fluctuations times 2.19 are transferred to the transmission.

So, in order to evaluate the effectiveness of the various dampening strategies, we need to measure the actual torque variation at the countershaft as it turns:
T(t) = Ratio*Torque(Omega*t) . Omega is the crankshaft's angular speed (rpm) at or near maximum torque, t is the time counting from a set point, Torque is the actual torque curve at the PTO shaft as it turns, and ratio is the total gear ratio between the crankshaft and the counter-shaft.

We need to look at T(t) for these cases
- no rear wheel cushioning, no engine torsion damper,
- rear wheel cushioning, and no engine torsion damper,
- no rear wheel cushioning, but a torsion damper fitted at PTO shaft or at the main shaft.

Because of numerous factors coming into play including rubber and chain non-linearities, the torque variation is best measured in an experimental set-up for each of the above cases. My assumption, which I cannot prove at the moment, is that torque peak values will be 0.4, 0.35 and 0.1 times the average counter-shaft torque value, respectively.
I very much doubt we will see figures for the second configuration at less than 0.30 .

-Knut
 
Last edited:
Because of numerous factors coming into play including rubber and chain non-linearities, the torque variation is best measured in an experimental set-up for each of the above cases. My assumption, which I cannot prove at the moment, is that torque peak values will be 0.4, 0.35 and 0.1 times the average counter-shaft torque value, respectively.
I very much doubt we will see figures for the second configuration at less than 0.30 .

-Knut

Thanks for the Reply Knut. I love the above answer. It smacks of open mindedness and an understanding that the best way to compare different configurations is to test them each of them and measure the results...
 
Status
Not open for further replies.
Back
Top