Summary – Presentation and reflection

We have reached the end of our project, and today we had the final presentation of our construction, and show what it was capable of.

The test in went through included different tasks. The first task was to cut the biggest piece of foam possible. It is shown on the picture below:

The biggest cut possible.

The biggest cut possible.

The second test was to cut out four large circles, that would fit into the big square from task 1. Our grous misunderstood that task, and cut four seperate circles instead of cutting the circles right on the square. But we still got some ciricles with a diameter of 17 cm anyway😉

One of the large circles.

One of the large circles.

The third task was to cut 3 squares in 10×10 and then 3 circles with the diameter of 10. That task was a little more difficult. We discovered that our belt on the x-axis didn’t work optimally and wasn’t able to pull the sleigh, when it reached the middle of the construction. Therefore, we tightened the belt a little bit more, and tried to cut another round of both squares and circles. That went way better.

The three circles.

The three circles.

The three squares.

The three squares.

The fourth task was to cut three gears of different sizes. After re-tightening the belt, this geomtery was cut without any problems.

The three gears of different sizes.

The three gears of different sizes.

The fifth task was to cut a random letter from the word MEKATRONIK. We got the letter K, and cut that too without any problem and with highly precision.

The letter "K"

The letter “K”

The sixth and last task was to cut any geometry, a so called free-style session. Since the name of our construction is inspired from a dinosaur, a Sauropod, we decided to cut a silhouette of such a dinosaur (See earlier post about our first dinosaur-cutting-session: https://m3k47r0n1k.wordpress.com/2014/06/01/our-creative-contribution/). Even though the first Dinosaur-cut turned out fine, we decided to try to make a geometry, which is a bit more complex. We found a silhouette of a T-Rex-cranium and traced it in Photoshop. Then we transfered the file to Inkscape, and then we were able to cut the cranium in foam. The result was very successfull and despite all the crooked lines, the construction was able to cut it with a high precision and accuracy.

The cranium in progress.

The cranium in progress.

The cranium up close.

The cranium up close.

All in all the presetation went well, but there’s still room for improvements. The fact that the belts were a little too lose, shows that this is an area we have to focus on for further work in order to optimize the precision of the geometries even more.

As a conclusion on our project, we think that we’ve accomplished a construction that fully lives up to our criterias. The construction is simple, consisting of few components and is easy to transport, seperate and assemble again if needed. Furthermore is the precision and accuracy high, and thereof, the results of the geometries are good.

Inkscape and GRBL

Inkscape is a program that is an open-source vector graphics editor, and similar to Adobe Illustrator. Yet, Inkscape is different from Adobe because of its use of Scalable Vector Graphics (SVG). SVG is an XML-based vector graphics format.

When we use Inkscape, we draw the geometry and the inlets and outlets. The geometry is being joined together by little marks along the edge. A shot of the screen is shown below. The green box is where the specifications are listed, and it’s here we adjust the speed if needed (the specification named “feed”). After many tests we figured that the longer and linear the distance of the different points on the geometry is, the lower do we need to put the feed. Otherwise, it will pull the wire too much and the geometry will be damaged (this is also mentioned in an earlier post: https://m3k47r0n1k.wordpress.com/2014/06/01/the-foam-cutter-a-full-description/)

10441965_10154187751570065_7659324958122103114_n

The GRBL controller is a software that will run on an Arduino. It’s designed to send G-code to CNC-machines. This is where we adjust the coordinates and the point of origin. The GRBL is suitable because it’s for three-axis-machines, X,Y and Z. It has also been tricky for us to figure out, how the axes should turn, because it often, all of a sudden, would mix the settings, and make it impossible to cut out a nice geometry. This would also affect the home-swithes. Why this error occurs can have mutliple explainations.   We tried to search for the problem, and found some possible reasons and solutions and this website: https://github.com/grbl/grbl/wiki/Frequently-Asked-Questions

10359169_10154187751580065_4875004810437379127_n

The foam-cutter – a full description

Our proces of building the foam-cutter has been a stabile process, where we have adjusted the design continuously.

We have some overall criterias and specifications for the final construction. We want to summarize them on this blog-post, to get a clear view on what our construction is capable of, and what advanteges it holds.

How does it work?: The construction is familiar to a so called T-bot, except from the fact that it has two individual belts. We achieve the linear lead in the x-and y-axis by having to sleighs made of acrylic material. The sleighs are fastened with the belts and goes along both the horisontal and vertical aluprofile. The first motors is fastened to the horisontal aluprofile and the second motor is fastened to the acrylic-sleighs. This leads the second sleighs up along the y-axis and leads the hanger and wire. The wire is fastened to the hanger by two aluminiumpipes and twisted around two screws. The hanger is made of tree and that’s how we avoid electric leading from the wire.

The plate, on which the foam-pieces will be cu,t is ply-wood and the foam is stuck to the plate by nails. The plate itself is being screwed on the aluminium-profiles.

The design: The criterias for the final construction are as follows:

1. The construction has to consist of a few components as possible (simple construction).

2. It has to be easy to seperate and collect again (the components are joined in a few places).

3. The weight should be low, which means that a man of average size will be able to carry it by himself.

4. The area, of which the foam is being cut, has to be maximized – of course in collaboration with the weight and simplicity.

Our final construction has a design that fully meets the criterias listed above. It’s able to cut big pieces, it has a weight of 6,9 kg (included the PSU and all other electronic components). It’s easy to seperate because it consists of few components and therefore has a simple design.

Materials: We wanted to mostly use standardcomponents in order to make it as easy for us as possible to obtain the parts, and also to make sure that the final construction wouldn’t be too expensive. We have used the 4 m. of aluminium, which was availiable from the beginning. The plate is made of ply-wood and the hanger is likewise out of tree. The components we had to make specially are the sleighs and some of the mountings. We made these out of acrylic material by lasercutting. All other components like screws, ball bearings and fittings are all standard.

Own experiments: Regarding the tests, we have made multiple. We had a lot of trouble figuring out how the GRBL-configurations of the geometries should be done. We tried several times to change the the speed alternating, depending on which geometry we were cutting. We eventually figured how we should fix the different files and that’s how we got our final results.  The pictures below illustrates how different results we got from changing specifications on different letters, such as the size and the speed.

In the end we figured that it was the belts that was a little too loose, and furthermore the gear on the motor was very loose and almost about to fall off, if it wasn’t for the belts that held it in place. After we tightened the gear, the running went way better for all the different geometries.

The letter A - we had some trouble with the speed of the running, and at a point, the wire stopped for a few seconds before it started running again. Why this happened is still a mystery to us.

The letter A – we had some trouble with the speed of the running, and at a point, the wire stopped for a few seconds before it started running again. Why this happened is still a mystery to us.

The fact that the wire holds still for a few seconds has a huge impact on the final form - some of the foam is melted away and makes holes in the geometry.

The fact that the wire holds still for a few seconds has a huge impact on the final form – some of the foam is melted away and makes holes in the geometry.

The letter E is smaller than the letter A, and it went way better. The wire didn't stop during the running, and the letter was cut in a nice shape.

The letter E is smaller than the letter A, and it went way better. The wire didn’t stop during the running, and the letter was cut in a nice shape.

Further work: If we had the opportunity for further work, then we would adjust different elements. First of all, we would fixate the hanger even more, to make sure that it wont vibrate when the construction is running. So far, the hanger can be affected when we run on a high speed, and start to vibrate. This affects our precision. If we improve the fixation, we can might prevent the vibrations.

Furthermore we might be able to optimize even more on the number of components and reduce it even more. Hereof will the weight also be minimized a bit more.

Tests with Inkscape

We have been testing some different shapes with the foam-cutter from Inkscape, and here are some of the very first results. We cut some circles and some gears. It went pretty well, and we figured that the optimal speed was 100. If it was slower, we risced to melt the the foam too much, and it was faster, it would pull the wire.

Some of our very first shapes cut in foam

Some of our very first shapes cut in foam