Pagename: science - Science
Category:Array
Post Count: 5Category:AI related
Pagename: science - Science

AI – steps used by LLMs

Leave a Reply

Your email address will not be published. Required fields are marked *

Category:3d printing
Pagename: science - Science

Drill bit gauge (3D printed) – 2mm to 5mm

It is a frequent requirement in many manufacturing workshops to put away random drill bits that have been used in various projects. Annoyingly, it is not always clear what size the drill bit is.  I often found myself measuring a bunch of drill bits with vernier calipers to confirm their size before putting them away in the correct spot. 

Often, it is something that can be done by eye.  Common sizes like 1/8″, 3mm & 4mm are easily identified when those are the only sizes used, but an issue arises for me when incremental drill bits are used. For example, 3.2, 3.3mm, 4.2mm etc. My confidence in estimating by eye is reduced somewhat and I end up reaching for the calipers. Annoyingly, calipers are useful for many tasks so they are not necessarily where the drills go and often end up going walkabout. Frustrating.

So, a gauge  of some sort to make the process quicker is one 3d printing project I had in mind for a while. It went through some iterations in my head until I decided to go with the super simple solution of using a thin block with slotted holes. It’s not very accurate but it is sufficient for the job. There is no real need to be more accurate than a tenth of a millimetre.

I did a few iterations which involved minor changes to markers and dimensions. I tied it to be piece of string to keep it in the correct location and have been using it successfully for a month or so now. It has been perfect for what I need. For the first few uses, I needed to confirm to myself that it worked as well as it does by checking against the verniers. Spot on every time.

It is faster than using calipers or the usual method of trying to put the drill bits into the holes in drill bit cases. Accuracy is low but adequate. As far as improvements go, I think it could handle more sizes and have clearer markings but it does what I need perfectly at present so I am leaving that version as is.

I uploaded the model to ‘printables.com‘ at: printables.com/model/428869-metric-drill-bit-diameter-gauge-for-2-to-5mm

Leave a Reply

Your email address will not be published. Required fields are marked *

Category:3d printing
Pagename: science - Science

Printing foldable objects

It was not long after I started printing some desired objects that I hit the snag of overhang. For example, I want to print a small box to contain some playing cards. Most of the structure is straight forward but an obstacle is the roof. No matter which way the box is orientated, an overhang exists. Of course, there is numerous ways to tackle this issue. One approach I am experimenting with is to print a part that can be folded into shape.

I did try googling the subject but I still did not get any clear parameters to use. Hence my need to do some experimentation.

In FreeCAD, I did a quick part using 3mm height and with a 90° wedge placed between two sides. The wedge is placed at 0.5mm above the bottom.

Here is the part in FreeCAD:

From my print, using PLA, I had a few observations…

  • Firstly, the 3mm height was not needed for structural strength I needed. I will be able to reduce that.
  • When bending into the wedge, I felt the plastic give way. I believe that means 0.5mm is too thick. Bending it after that point seems ok. I can bend it repeatedly without it splitting in any noticeable way.
  • When bending into the wedge, the part wants to bounce back. Hence it seems that 90º is not enough to allow a 90° bend. Interestingly, bending the part in the opposite direction was fine and it would hold a 90º bend. Of course, a chamfer appears when bent in this direction.

With the next few iterations, the main problem seems to be with my FreeCAD skills or maybe a glitch in the program itself. I created the wedges for the corners by inserting a segment of a cylinder and positioning the cylinder so subtracting it from the model resulted in the required wedge. However, while this mostly works, each time I have printed the result I have had a remaining wedge left in place. That is, it was not subtracted from the model. Apart from this annoyance, the next few iterations were improvements. Changes were:

  • Adding the missing walls and the bottom to make five sides of a cuboid.
  • Reducing the wall thickness from 3mm to 1mm. 3mm was overkill for this application.
  • Increasing the size of the wedge from 90° to 92°. This did indeed improve the bends but I need to improve the model to be sure it is the best angle to work with.

Here is the 2nd iteration. A few obvious mistakes can be seen including incorrect length on the side and the presence of the wedge I described above.

 

 

 

This was the third attempt:

I added some filament from a 3d printer pen to hold it in position. Mostly this was fine except for the remaining wedge issue again. A deck of cards fits in nicely.

Next iteration finally got the cuts/wedges in the right orientation. This version was largely as desired but I had reduced the wedge angle to 91º and I find the corners still want to bounce out a little. It is acceptable but I think going 92º produced the best results.

Here is the print:

This was the desired product. Now to add some features to make it a superior container for a deck of cards. 

Leave a Reply

Your email address will not be published. Required fields are marked *

Category:3d printing
Pagename: science - Science

Cool 3d prints

Here is a list of 3d prints that I want to print when I can. I have a focus more on tools on puzzles, and less on decorations or figurines.

Tangram: A puzzle I used to have as a kid which I would like to try again. This print seems similar to how I would want it but I think it would be better to just throw it up in FreeCAD so I can get it how I want.

 

Leave a Reply

Your email address will not be published. Required fields are marked *

Category:General
Pagename: science - Science

Third twenty in twenty – Mathematics is worthwhile

Personally, I love a good puzzle to work on. So, for that reason, it makes sense I enjoy mathematics more than the average Joe. Still, even with that in mind, I am perplexed at the resistance applied to avoid learning all things mathematical. This topic comes up fairly regularly now I have young children of school age. There disinterest is accentuated by the fact that neither has a definite idea on what they would like to do as a future occupation. I certainly don’t won’t to hit them with the ‘well those occupations require mathematics’ line for fear of encouraging them to avoid those particular occupations. Both kids have resented the mental effort required to learn mathematics at various points. Some of my children’s teachers have even appeared to acquiesce with the students in their dislike of mathematics. Very disappointing.

One worn out argument that is often aimed at mathematics is ‘I will never use it, so what’s the point?’. Indeed, on face value, the argument has value. There are not too many jobs that require employees to break out their pen and paper and start throwing Pythagoras theorem around. Yet, many of those same workers will apply mathematics in an implicit sense. Making sense of dates and times is a frequent task for most. Being able to make adjustments to various measures of magnitude is a given in daily life. Got enough money to buy the milk? How many extra eggs to make that double batch of pancakes? This is not even touching on the more advanced ways we can and do use mathematical thinking, generally without even relating it to the mathematical skills we are actually using. The same skills we learn at school.

Leave a Reply

Your email address will not be published. Required fields are marked *