[Diy_efi] RE: [offlist] Strain gauge on engine mounts

[niche at iinet.net.au]

At 01:55 AM 21/04/05, you wrote:
>Sorry for not being clear. Yes, that's what I was talking about. If possible, all you'd need to do is remove the driveshaft and apply a known torque and measure.  It probably twists more than you think across that distance.

The thing is I have little idea just how much twist, however it wouldnt be hard to change the
driveshaft for one which did twist a fair number of degrees for a known torque range,
rather than worrying too much about whats there it wouldnt be hard to change it,
still much cheaper I'd guess than the alternative...

>However, I don't think it's very critical on the pitch - almost any "convenient" number of teeth should work fine.  The only consideration is whether the applied torque *changed* between the sampling times of ring gear 1 and ring gear 2. Since the torque is relatively constant (certainly over the time each tooth goes by), and the entire ring gear is phase-shifted the same, it should be quite accurate. Again, I think the key to success for this is sensor coupling, dealing with run-out/wobble/movement of the driveshaft. Plus any movement of the sensors relative to each other would introduce potentially major error - that's one disadvantage of a longer distance.

I'm still favouring the split driveshaft type, ie the last piece of my shaft which goes
to differential would be way too hard, I like the horizontal section from auto tranny to
center bearing, one could make up a floating cage bound by the tranny support
so it reduces relative effects from body near the center bearing...

>(Hmm...maybe a laser pointed at teeth and detecting the reflection/back-scatter, or even precisely painted bars on the driveshaft itself?)

Now thats even more interesting, wish I'd thought of it ~`:o)

You can get these silver reflective strips that come with optical
tachos, then rig up a pair of keychain type led laser modules to shine the
light, sensors are cheap and could be easily collimated if need be,
any wobble is likely to not be differential in nature and any that
is would likely be filtered out. Of course one would have a processor
taking samples, discarding outlyers and applying some low pass
filtering, neat, could be cheap and fairly easy to install I guess,

I have vision of large wide multisegment linear led display across whole driver
side dash streaming from left to right as one accelerates or pulls a caravan,
I think the tractor pullers could even get telemetry and wow the crowds with
a huge torque competition prize and large vertical segment displays at night !

<chuckle>


Rgds

Mike





>-Marc
>
>
>At 1:18 AM +0800 4/21/05, niche at iinet.net.au wrote:
>>mmm,
>>
>>I'm not quite sure I understand what u think I am getting at,
>>
>>I have a two part drive shaft, the bit from the end of automatic
>>to center yoke is pretty well placed in transmission tunnel and
>>doenst move around much. So I was thinking placing a ring
>>gear on each end rigidly fixed to the drive shaft, the only piece
>>of 'stuff' between the two ring gears is the driveshaft, no coupling,
>>no rubber just hollow metal tube. Sure there wont be much twist
>>of this tube at torque but there will be some, and what I'd like to
>>know is just how much...so I can derive a ring gear pitch,
>>
>>Rgds
>>
>>mike
>>
>>
>>
>>
>>
>>
>>At 12:33 AM 21/04/05, you wrote:
>> >One good thing about using rubber is it's easy to "calibrate" on the bench - just put a known torque on it and measure the twist. Not super-accurate, but probably get within a few % of actual. Can also correlate the torque curve of this to a real chassis dyno. I do think the same thing can be done replacing the rubber with the driveshaft. In any case, the only real challenge is getting the sensors rigidly mounted, and to where they float with any shaft/coupler movement.
>
>-- 
>Marc Reviel
>
>PowerLogix
>http://www.powerlogix.com
>_______________________________________________
>Diy_efi mailing list
>Diy_efi at diy-efi.org
>http://lists.diy-efi.org/mailman/listinfo/diy_efi

Regards from


Mike Massen
Perth, Western Australia
VL Commodore Fuse Rail that wont warp or melt !
http://niche.iinet.net.au