The Raupe project.

Raupe is a german word. Normally, english words are appointed to projects, but that's getting boring. So I chose to give this project a name that reflects the origin of the vehicle (Austria) and the country of our friend Fiffi (Germany) who arranged these kits for us.

Apart from this, my family decided to start a naming contest for Raupe. They came up with the name 'Speedy'. What else, for a crawler vehicle... I haven't yet given in to use that name. Perhaps it will be best to just ignore it.

Building Raupe.

The Raupe project is about building a small robot in the form of a 'tracked vehicle' similar to a big Caterpillar bulldozer or an armoured tank. We start this project on the base of the 'Raupen Fahrzeug' as made by Winkler Schulbedarf in Austria. For an indication of the kit and how it looks like, please pay a visit to their website or click on this link for a picture of the completed kit.

Or, if you're impatient, just take a look at the right ---->

To the right, you see how I mounted the wheels. This is not the way Winkler instructed things. No, I was stubborn again.
I decided to change some things:

What you see here is the motor-side of the chassis.

In this picture you see my modifications to the original design so far:

This picture is just the bottomside of the current Raupe. You see the holes for the sensor array and the LED's in the 'windshield'.
In the back, between the wheels, you see a piece of bent metal (cut from a piece of metal that used to be in a PC cabinet in front of a yet unused 5,25" drive bay) that acts as a retainer for the motor carrier plate.

This is a close up of one drive wheel. As you can see, I cut these wheels from a sheet of plywood (the 12 mm part) and a piece of Eucalyptus timber-panel (the 18 mm part).
Both pieces of wood were cut with a 'seven holes saw', set on a diameter of 1.75" (45 mm), in a drill press.
The gap between the two pieces is for guiding the rubber fins on the inside of the tracks. The composition of the drive wheels is as follows:

Within seconds, the glue settles and the wheels are permanently connected with the shafts. So be careful: you only get ONE chance of mounting the wheels after you applied the glue on the shafts...

Here you see how the plug is inserted into the (wide) disc. You can push the plug in over 30 mm. The last 5 mm need to be hammered in since the plug has some fins on the side. Don't cut off these fins. They will give the wheel extra driving power.

What you see here is a kind of exploded view of the wheel construction. On top of the motors is the box with the plugs (brand and type well visible).
In the center you see how the parts are combined. The plug slides into the axle-hole of the wooden discs. Between the two wooden discs you still need to put two washers for making space.

The carrier wheels are loosely coupled to the axle (a 6 mm threaded bar?) with a lot of nuts on it to fix the wheels in place, but such that they can rotate freely.
These carrier wheels are smaller diamater wheels (approximately 30 mm diameter) but cut in the same way.

On the right you see how the carrier wheels are constructed. The photo should say more than a thousand words, so I stop writing here.

A picture of my 'wheel-yard'. OK, when you're sawing discs, you should make enough of them for several tests.

On the right you see how the tracks will most likely be mounted. I haven't figured it out completely yet. There are some factors involved here and I must see to meet all demands as close as possible.

I need to take care of the following:

Let's see how far we come here.

Well: this far! As you see, Raupe is functional! I connected a 1.5 Volt D-cell (the big one) across both motors and it RUNS!
Both motors are virtually in sync. At 1.5 Volts, the speed is roughly 10 cm per second, which is just about ideal. And both motors rev at the same speed since Raupe doesn't tend to go off a straight line.

Warning! Red Alert! Remote controls on the loose!

Worse: a PHILIPS remote control is crossing the road. Raupe is going like crazy and since both motors are connected via massive wires it cannot steer so there's only one way to deal with this RC:


On the way back, you see what damage Raupe did to the RC. Not much. I was already glad that the RC didn't do damage to Raupe!

As you see, the tracks are steady on the wheels. I was a bit afraid that the tracks might have a tendency to roll themselves off the drive wheels. Until now, this has not happened.

I also took Raupe for a walk. It didn't like to be outside. Too windy. It might catch a cold, like it is now.
Perhaps later, but I reckon Raupe will mainly spend it's life indoors.

Never flip over a turtle! Or a sheep! They will die in due time since they cannot roll back by their own.

Same goes for Raupe: it has no SRM (Self Righting Mechanism), so if flipped over, it will stay like that. Unless the Master (aka Brilsmurf) is around. Lucky for Raupe that I turned it over.

When flipped over we can see the way the wheels are mounted to the chassis. These pictures are very difficult to expose with a digital camera due to the extreme contrast differences between the dull black tracks and the almost white birchwood panels.
Nevertheless, we see several things here:

One more picture of the flipped over Raupe vehicle. Use both pictures to extract the amount of information or detail you need.

Here is a close-up of the clothes-peg axle mount:

A top-view of Raupe. Not much to tell, to be honest. You see the way the motors are wired to the D-cell. YEARS ago I bought some of these D-cell clips. The year 2004 proved my aquisition true...

Back to driving!

See raupe tumble and fumble it's way through the livingroom. Less than a year ago, a young Belgian genius slept here on his airbed and now it's the playground of a tracked vehicle.

You can see that Raupe has good taste. It just worked itself over a Slackware Linux textbook. :o)

This picture is important. Not for the Slackware book, but for the shape of the track. You see that the track is concave on the trailing side. When riding forward, the motor pulls the tracks from underneath the vehicle, thereby pushing the tracks up on the trailing (top)side.
The solution here is tensioning. The tracks need to be tight on the wheels. Slacking tracks cause wheel-slip. For this reason I elongated the round holes for the front axle fixation. Now I can slide the loosened bolts forward and backward.

Another problem is related: when driving backwards, the tracks are pushed underneath the vehicle. And when the track tension is too low, this will lead to slipping of the track on the drive wheel. Therefore, a very good tension is needed for driving backwards.

Ever heard of the 'Domino principle'? If not, ask your history teacher. (S)he will give you a lecture or two. Anyway, this is Raupe's domino principle: if the hurdle is too high, and the battery is not in the front section, Raupe will stand on it's hind legs until it tips over.

Back to track tension. I solved the first track tension problem by elongating the mounting holes for the front axle.
This worked well enough for forward driving. Drving backwards still is a bit tricky. If I move the front axle more forward I will get more tension on the tracks, but the motors also need more power to drive it all. So a better tensioning mechanism must be found.

Probably this is why Winkler use their 'flat track' mechanism like it is used by most 'track propelled excavators'.
On 'flat tracks' the tensioning is easier.

Today, I spent part of the day thinking about how to solve the tension problem. This was wrong. The tension is ONLY a problem when we go for friction drive, i.e. we rely on friction between drive wheels and rubber tracks to propel the vehicle.

This is putting the horse behind the carriage. The problem was not 'How to improve friction'. No, the REAL problem was 'How to propel the tank'.
In essence we have two mechanisms at our proposal:

Some observations of Raupe: The 13 cm circumference is nice. If I apply 13 pins along the circumference of the wheel, I get 100% drive. But I don't want to do that. The circumference is slightly over 13 cm, so sooner or later the drive pins don't mate anymore with the trackteeth. An unwanted situation.
So I decided to go for partial full drive. I will put two driving pins along the circumference of the two drive wheels, 4 cm apart. That will give me around 40 degrees of friction drive and 320 degrees of sprocket drive. In case the sprockets and teeth go out of sync, the fraction of frictiondrive will easily accomodate for that.

So I took off the tracks and looked how to modify the wheels from frictiondrive to sprocket drive. I decided to mill away two shallow grooves in each drivewheel.

You see the result in the adjacent picture:

The milling was done with a small electric drill and a ball-shaped bit. The groove must be approximately 2 mmm wide and at least 2 mm deep. For the pin, I chose a small nail, 1.6 mm diameter. I cut four of them to size and next came the difficult question: how to fix the pins in the wheel?

To cut a long story short: there just ain't no other ways to fix two such diverse materials than using hot-melt glue.
'Normal' glue either will not stick good enough to the wood, or it will take days to cure. Also, most traditional glues do not fill cavities. Hotmelt has none of these disadvantages. It will fill cavities. It is strong and as long as it is fluid, it will accept materials and the parts can be corrected.

When glueing things together, I decide to also inject some glue into the gap between the two wheelsegments. This adds stability to the wheels and it reinforces the driving capacity in frictiondrive phase.
On the right, you see the pin, fixed into position. Each wheel has two of these pins.

Driving time again!

I made some barricades and went to the garden table with my li'l friend. It's a wonder to behold: forward and backward driving are now a breeze. Raupe takes on all barricades with the same ease, no matter if we're going forward or reverse.

For us, tinkers, it is essential to be able to see what we're doing. So sometimes we need to put marks on tracks. Which is not very easy.

I tried to do it with stickers. It looked to work, until I found out one fell off the track... The rubber is a special kind with very low sticking power for normal adhesives.
So I took my can of acrylic paint (do NOT use alkyd paint due to the solvents in that kind of paint!) and put some blue marks on the tracks. This worked fine and it showed that the left motor is NOT equally fast as the right motor. The left motor runs about 15% slower, so Raupe tends to make left turns on longer runs.

More to come....

Page created August 2003,