Making Machining Greener

So it's fair to say machining isn't always the greenest thing in the world so I try to make every little effort I can.. Mainly reusing materials as much as possible etc. One thing that used to bother me was the amount of wd40 type stuff I used to get through in the shed.. It's useful stuff for freeing seized stuff, penetrating and loosening rust, cleaning parts and also blowing swarf out of holes etc... However its expensive, uses hideous propellant's which I'm a bit sensitive too and definitely not green. So the solution! Well I've long known that diesel is an amazing penetrant so a mix of a few hundred ml of diesel and then a few drops of engine oil creates a mix that in my opinion is as functional as wd40 and I even occasionally use it as a cutting fluid. All you need to do is find a hand pumped spray bottle... I favour this princess hair detangler one as the vibrant pink is easy to spot in the clutter!

Materials play... CNC routing cardboard

So just had a very quick 15 minute CNC session as time is a bit short at the moment! I've been meaning to see how it is to rout cardboard (CNC router as poor mans laser). After a quick play its OK.. Nice to be able to cut a material in one pass and pretty darn quickly in terms of feed speed. I was worried that it may tear material rather than cut but this wasn't the case and even the smaller ring which is only 3mm wide didn't distort when cut. So it's good.. However it does have the disadvantage of leaving a small burred edge which you don't get with a laser. Also, obviously, your sacrificial material has to be something that can be cut/touched at the same keep and of feed rates or else you may break tools! So I fancy when time allows playing with some 3d print slicing software to see if I can slice a 3d model into layers/contours and then cut from card..

Pocketqube Chassis Prototyping

Lately I've been considering Pocketqube chassis development and got to the point of prototyping a few ideas and working out a few engineering operation sequences. The first challenge began with trying to find 6061T6 stock that is 50mm rather than 2 inches (come on world it's 2016....!!)

I sneaked a bit of time on a horizontal bandsaw to cut some oversize blanks then used my small manual mill with the biggest endmill I own (16mm four flute with 63mm cutting area) to bring the blank to 50mm length. It would be preferable to face mill the piece to size but its easier to use the side walls of the stock piece as a datum.

Then over to my CNC router to play with some skeletonising ideas which was an interesting set of procedures as up until this I have tended to have only done jobs requiring the job to be cut out of sheet material. This means usually I only have to worry about zeroing the Z axis and pay little concern for accurate placement of the stock material. Using the Tee slot edges and a piece of angle stock and a small parallel it is easy to place the work piece accurately in line with the X and Y axis and touch on using feeler gauges (still haven't finalised my conductive probe!) and using a G92 command to set the corresponding axis position (minus half the diameter of the slot drill in the case of the X and Y axis).

What I would really like to do next is model some of the chassis ideas/designs I have in OpenScad and then try and perform a Finite Element Analysis (FEA) to determine which design should be most resilient to the resonance profile of a rocket launch... but despite reading a fair bit in the last few days it may be some time before I get my head round FEA techniques!!

Next challenge appears to be finding 6061/2 in 5 or 6 mm hexagonal bar stock ... but not 1/4"!

Knobs and Power!

So a couple of little projects in between everything else this weekend. Firstly a classic hack turning an old ATX power supply into a bench top power supply.. I'm not going to go into too much detail as there are THOUSANDS of instructables and blogposts on how to do this out there.. I had to add a 10 watt 10 ohm resistor dummy load across the 5v rail to make all the outputs create the correct voltages. Of course I used the CNC to make a little breakout panel which was interesting in terms of I realised I am at the point where it is quicker and more accurate for me to whip up a quick drawing in inkscape and rout it rather than breaking out a drill and saw!

Second up links to my previous post .. all the universal clamps on my stand system are closed by 4mm threaded systems usually a bolt with a handled nut like so.. I hadn't made any more of these and currently am sharing the couple I'd made between different bits of the system. These are great but equally take a bit of time to make as they require quite a few operations. The handle parts are press fit into a 4mm reamed hole as well as the tapping etc..and they also use up my meagre supplies of nice EN1A steel! So I whipped up some nice knobs from some 6mm MDF with a routed hexagonal hole into which I could press fit a M4 nut with a blob of super glue, they work really well.

Projects within Projects!

Got some time in on some projects of late and today I finished making a small project.. that I needed to make as part of another ongoing project! The overlying project requires some tapping of M2 threads through some 10mm aluminium bars and I am keen to make the threads as accurate as possible as I want to make sure as much of the thread is engaged as possible when the bolts are fitted. I am reasonably good at hand tapping and also use the tap in a pillar drill or tap in the tailstock technique but for really small taps I find that you lose a  lot of feel this way which can result in broken taps or malformed threads. So I needed some form of sensitive tapping stand that allows you to keep the sensitive feel whilst guaranteeing the vertical introduction of the tap to the workpiece. As this is only for small work I decided to make an arm and chuck shaft that would fit onto my existing (and ever growing) clamp stand system that started out based on Harold Halls Dial test indicator stand. So the chuck shaft is held in the bush and arm vertically whilst still being able to be turned and lowered and raised through the reamed brass bush. I made a bit of an error on the hole to attach the assembly to the main upright and made the hole slightly oversize (dodgy sharpened drillbit!) Although the clamp would clamp it only just held so I added 2 grub screws which hold it firmly. It's actually quite good that way as I can roughly set the clamp height quickly and then secure it afterwards.

The chuck shaft is formed from a bit of mild steel and a cheap small drill chuck from ebay. I turned the shaft of the threaded keyless chuck into a taper and made an undersize taper bore into the steel shaft, then a careful few pressings on the arbour press and it is well attached and straight. I just need to make a handle section to apply the required turning..I'm considering whether to make a press fit knob for the shaft or whether to machine some flats into it allowing tapping handles to engage it. 

Of course the thing about all the accessories being compatible is this can be used in conjunction with other items! It also occurred to me the chuck shaft could be clamped in one of my universal clamp assemblies and used as a pin vice. 

Keep It Simple (Stupid) DIY Rocket Altimeter SD Datalogger project

Been working on this little project, a while back I prototyped a diy altimeter using an arduino pro mini (8mhz 3.3v) and a BMP180 barometric pressure and temperature sensor. I have now replicated the working breadboard prototype and made this tiny point to point/deadbug style version. 

The reason I have aimed to build it as small as possible is that I wanted the altimeter to fit inside the payload section of this small "Star Stryker" Estes Rocket kit I built a while back. This rocket payload section probably represents the smallest payload capable rocket on the market so the design will therefore theoretically fit into any payload section.

Weight wise it is 6.5g without the battery and currently 13.7g with a battery.. however the 200ma lipo could be replaced with something smaller (it was just what I had lying around!)

The device is activated by the switch that means that the payload section must have an access hole/port to enable the switch to be toggled with a tool.. but this is ok as the BMP180 sensor needs the payload section to be vented anyway to equalise the pressure and take an accurate reading. One slight addition I am going to make is to attach a larger LED to the pro mini. This LED will be lit on boot and will remain lit during a delayed setup part of the code (10 seconds or whatever is needed) to allow time for the operator to return to the launch controller. Then, when the LED is off, the main loop of the program is initiated writing data to the sd card at 20hz

The code currently is a bit shabby (I am no great coder!) and is just a mashup of  SD library and a BMP180 library sketches.. it's crude but it works! I'm going to work on the code with a good friend in the new year and when we have it polished (possibly dumping data to a .csv file for quick plotting) I'll opensource the project in case it's useful to others! 

OpenRocket and rocketry CNC workflow...

I've been playing with OpenRocket for a while now, essentially this little free java application allows you to design rockets and simulate their performance with an extensive database of rocket engines built in. It calculates both centre of gravity and centre of pressure and uses the Barrowman equations to calculate the stability (or lack of) in your design and has lots of aerodynamics code built in calculating drag coefficients etc. Being able to simulate launches and recovery is excellent and allows you to both optimise your design in terms of feasible altitude but also can allow you to fly within the limits of a particular altitude ceiling. Simulation also is important when you get up to qualifying certification flights for high power rocketry as your rocket needs to perform similarly in real life to what you have submitted in simulation data.

So all of this was exciting as I start to design my first scratch built rocket attempt, however an added excellent bonus is that Openrocket exports it's simulation data into a PDF document which I hadn't done until recently as I wanted to printout some data. The export function also will print out various other useful items from your design including fin templates and centring rings (rings that hold the inner engine tubes to the outer airframe tube) as PDF is a vector format it is readily accepted into most CAD/CAM environments and therefore it is a trivial matter to create tool paths to cut fins and rings etc on my CNC router. Slick!