How Can We Improve Odometer Accuracy for Precision Testing in Civil Engineering?

[AshNZ]

Hi,

I'm a civil engineering student in a summer internship at the moment but I've come across a mechanical issue. We want to buy or build a very accurate (+- 100mm) odometer for a large truck. We have a Survmaster precision odometer in there but it's only accurate to +- 1.0m which isn't good enough.

Do you have any ideas about any devices that we could build? It should be readable from the truck cab and driver should not have to come out to adjust too many things (in other words it should not be time consuming)

edit: Sorry should have mentioned it before, we're looking at that accuracy over very small distances (2-3m) and at a very slow speed. We need to be able to measure 2.7m +- .1 m.

Any ideas??

Thanks

 

[NobodySpecial]

Over what sort of distance?
You are going to have a real problem with a rubber wheel on a real road not slipping or deforming with that sort of accuracy over any resonable distance

 

[russ_watters]

Heck, just making a turn will cause the distance traveled by the wheels to be different from each other!

 

[AshNZ]

Sorry should have mentioned it before, we're looking at that accuracy over very small distances (2-3m) and at a very slow speed. We need to be able to measure 2.7m +- .1 m.

 

[Jobrag]

I can't think of a process that would require moving a large truck with such accuracy over such a short distance, perhaps you need to rethink the whole process and remove that requirement. If that is not possible something as simple as a rule laid on the ground with a pointer attached to the truck would be cheap and simple.

 

[NobodySpecial]

Can you rent an RTK gps or total station theodolite ?
Or just a laser distance meter?

 

[Topher925]

A good ultrasonic distance meter should be able to handle this no problem. One of those stereo laser distance thinga-majigers would probably work well too.

 

[Mugambo]

You might want to find a super flat stretch of 30 meters (length of truck + 5 meters approx) and use the above mentioned laser/ultrasonic distance meter. Rinse and repeat

 

[AshNZ]

Thanks for your replies. Seems like I wasn't 100% clear. What we basically want the truck driver to do is sit in a parked truck, roll the truck forward slowly and stop when he's driven 2.7m. This probably rules the laser distance meters out. The truck isn't an 18 wheeler. It's just medium sized truck with no trailer.

We were looking at having http://www.bing.com/images/search?q=distance+wheel&FORM=BIFD#" and attach it to the truck and then somehow connect it so the driver can read it from the truck cab (maybe with a camera?)

It's not a one off thing that we're doing.. it's part of R&D that my company is undertaking.

 

[Mugambo]

Tape a novelty laser pointer facing down to the freehub of one of the front wheels. Print scale markings for about 5 meters using a drafter (at 1:1 ratio of course). Lay it down on a flat stretch. Align laser pointer to the printed scale. You should be able to calibrate it within 5 trials.

 

[NobodySpecial]

A tacho solution would be fairly easy to implement - on a truck the rear axles are usually accesible, it would be easy to mount an angle encoder on them and either measure the wheel cirumference or just drive and calibrate it.
But you will have accuracy issues with load and surface (the tire will compress changing the circumfernce) or if the truck is turning.

An absolute system with some distance measuring machine and an external target is easiest, assuming you are always moving over the same course.
If not then the easiest solution (assuming you are on a road) would be a trailing wheel - like a survey wheel - but with a fraction of a revolution counter. Just a slotted disc a photodiode and a micro-controller if you want to do it yourself.

 

[russ_watters]

Tape a novelty laser pointer facing down to the freehub of one of the front wheels. Print scale markings for about 5 meters using a drafter (at 1:1 ratio of course). Lay it down on a flat stretch. Align laser pointer to the printed scale. You should be able to calibrate it within 5 trials.

...and install a webcam slaved to the laser pointer and outputting to a laptop in the cab.

 

[Jobrag]

Is this just an exercise or is there a practical application?

 

[AshNZ]

Thanks for the suggestions. We close for Christmas today so will come back to this in the new year.

Jobrag, this has a practical application while testing pavements. We have a proprietary test similar to Benkelman Beam test and we're just working on it.