Monday, 15 May 2017

Bracketless Pocketqube Chassis Concept

I've been meaning to post about this since I went to Delft to talk about it but only got round to it now, despite the fact that the chassis I made to demonstrate this idea (pictured) currently sits in Switzerland en route to travel to Kathmandu this weekend to hopefully inspire engineers in the Orion Space program.

So I've been asked to machine a few Pocketqube chassis from square tubular stock, lots of people in the community seem to favour a tubular chassis approach as it has lots of strength and it has some shielding and thermal benefits over PCB construction. Invariably being asked to make skeletonised tube chassis  has led to also being asked to make tiny brackets that will attach end plates to it. Now ... it's certainly possible to do this but it takes a lot of time and may not be too accurate. Time wise it is tricky because .. well.. its fiddly... work holding such tiny items is a pain and when multiplied by a minimum of 8 brackets it loses all sense of fun! The other issue is they are difficult to make accurately.. so for the uninitiated in metal working, when you buy a square rod of say 6mm aluminium its surface finish will be OK for most things.. but for assembling something critical like a satellite end plate to the chassis it needs to have a good surface finish so that the bracket face sits flush onto the interior wall or walls of the chassis tube. This means that really for an 6mm bracket (L shaped or a 3 wall/surface design or just a cube with tapped holes) you really need to machine each of the mating faces which means you need to start off with probably 7/8mm stock and do a lot more operations. Finally the other consideration though is that the internal finish of the chassis will not be very precise either and will require spot facing (a small machining pass that levels the surface finish) but as this will be on an internal face (and usually in a corner) machining is often not an option.

This can lead to inaccuracies such as the bracket not sitting flush into the corner as shown (obviously not quite as dramatic as this in real life!) such as in the image below. This issue will reduce strength and particularly will cause issues when we consider that a M2 bolt with a 0.4mm pitch threaded into this bracket may only have 3 or 4 turns of thread to go into and at an angle the thread contact area will be reduced weakening the clamping effect and increasing forces onto the chassis and end plate.
So to overcome some of these problems I've developed an idea for a bracket less PQ chassis, simpler to machine, possibly stronger and also certainly more flexible in terms of mount points for internal or external components. My first prototype also adheres to the PQ60 specification. 

So basically the chassis is a tube design mounted to the base plate with a stack of PQ60 PCB mounted through the tube on threaded bars. Into either end of the tube a 1mm escutcheon is milled to a depth of 1.6mm allowing each end of the tube to receive (a light press fit) an FR4/PCB end plate. 

8 standoffs are made to be attached to the ends of threaded bars and these receive the bolts through the larger end plates. The making of these standoffs is much easier than brackets as they only require the mating face to be machined/cut square and to a reasonable surface finish, in fact if you drill the hole through the stock first I used a very fine slitting saw in the milling machine that cut these to a satisfactory finish with no further operations.
The stack assembly gives a benefit that the 2 end plates are pulled into the structure and (whilst I need to test on a vibration table) intuitively this seems very strong. 

Essentially as the internal stack floats in the tube this approach also maximises what can be done to the chassis tube, although in my prototype the skeletonisation is minimal any point of the structure (apart from the area surrounding the end plate) can indeed be skeletonised or used as a mount point. This is also true of the base plate as now fixings can be made at any point in the base plate creating maximum flexibility of the structural design. 
As ever, feel free to use any of these ideas and or get in touch if you need a chassis construction making. 

No comments: