Showing posts with label Engineering. Show all posts
Showing posts with label Engineering. Show all posts

Saturday, 24 January 2015

SLA 3D Printer Project Log 12: Dropping the Hammer on lack of construction

You know that feeling when you've finally been liberated to make real progress on something?  In a sea of drudgery and busy-bodying in the real world, this is a shot of Dopamine straight to the brain.  This is even more so when you've been procrastinating something for so long that spiders the size of your big toe have set up ambush positions in your cooking utensils (as happened to me when I cleared out my shed; yes, my spine is Petri Jelly, no need to remind me...  *sigh*).

Nails, the classic solution for design ineptitude - that or excessive amounts of glue and filler!

The rush of finally taking action is, as we all know, the best time to find out we've been imbeciles – as I found out when I discovered my top vent had no tabs for locating the parts together!  I had basically expected the builder to be able to glue these together against gravity no matter which way up the join was being made.  Rest assured this will be fixed in the final 1.0 release, but for now, we can get busy with a Pin Vice (Check out the review here!).

If you think this is laborious, try correcting a hammer drill gone haywire...

As I examined the Z carriage to which the Build Bed is mounted to, I couldn't help but notice that there was a slight warp, a horrifying scenario for a part absolutely dependent on being straight and true!  At risk of making the Z carriage too heavy for the Stepper Motors, I knocked together a barn-door solution on a whim...  Unprofessionalism?  What is this, the Intel FABs?

Spare Extrusions fixed in place with Stud-Locked Nuts and Bolts; a nice solution, though I'd be lying if I said I wasn't concerned for strain on the Z motors.

Among the bits that I bought in a project-buying frenzy were some black, chunky tapes with a faint magnetic field.  These are for mounting the UV-Reactive Acrylic Window Panels.  TBH I'm a little worried about their powers of retention given that most ferrous things are barely attracted to them, but at least my Hard Disks are safe :)

Pro-Tip for those doing this:  Dampen the plywood surface where the ends of the tape meet, apply cyano, then affix.  Safeguard against peeling ;)

One final note is how we can expect the panels to be mounted; I've found some Slim T-Nuts for the extrusions that don't need to be slid in before the panel is aligned, so I can mount the panels after the frame is built – you can't imagine how much rage this will save!  On the downside, I've missed out on the chance to become the next Francis...

This thing locks against the extrusion when tightened, so no worries about it spinning in situ.

Locking in action, shouldn't too hard to remove them again.

Stay Productive...

Friday, 4 April 2014

SLA 3D Printer Project Log 5: 200 Step Program

I've recently obtained a Raspberry Pi (and instantly overclocked it - reached 900MHz CPU, 500MHz GPU and 600MHz RAM with an overvolt of 8), as skeptical as I was at first, this £25 Set-Top Box with pretensions to being a Linux PC is giving me some real inroads in this project.

I've already got it to run a NEMA 17 Stepper Motor through an EasyDriver V4.4, and my coding prowess is as extensive as the Waterloo and City Line; I did this via WiringPi, a library for the Pi that enables you to use the GPIO Pins without using Python or typing sudo (super-user do) ahead of every command.

The commands entered were:
  • gpio mode 1 pwm - this assigns Pin 1 to PWM operation
  • gpio pwm-bal - this enables Balanced PWM, which is SoC default
  • gpio pwmr 100 - this sets the range to 100, or 0.1ms per unit
  • gpio pwm 1 - this sets the duty cycle to 1, or 0.1ms out of 10ms
Keep in mind it is possible I am hideously wrong at this stage, I don't own an oscilloscope and with that in mind I am a cretin.  I frequently used PicoScope in Electronics class in College so that'll be my first port of call.  I believe I also need to vary the frequency to change the speed, not the duty cycle; maybe that'll need a dedicated board...

In other news I have finished my Projector Mount for the SLA 3D Printer, images below:
Frontal View, the width of the frame (sans handles) is 580mm.
Isometric View.
Rear view, showing the main frame.
A side-on shot showing the guts of the traverser.
I'm slightly doubtful about the 8mm steel rods I'm using to support a Projector which may weigh up to 5kg, so let's do a few Beam Bending Equations :P

Let's assume that the projector is positioned dead centre of the supporting rods in both axes, and that it weighs 5kg, it's centre of mass is spot-on in the middle of the Traverse Plate and 100mm above it (to account for an early, heavy and highly bulky DLP projector).  These are the crucial dimensions (for moment calculations):

  • Horizontal Rods are 560mm long, so 280mm is the moment arm
  • Vertical Rods are 565.5mm long, gives a moment arm of 282.75mm
  • There is 62mm from the Horizontal Rods to the Traverse Plate
  • This falls to 13mm with respect to the Vertical Rods
  • On the Traverse Plate there is 170mm between the centres of the bearings (for the vertical axis)
  • 317mm is the distance between the Horizontal Bearings
In order to find the deflection (which will cause uneven projection distance and hence ruin our prints), we need to find the moment arms and hence the force experienced by the beams at their various points; we know the Mass of the Projector is 5kg, this multiplied by 9.81m/s2 gives us 49.05N of force, which for the sake of simplicity we will disregard for the vertical axis for now, and focus on it's effect on the horizontal axis.

Mooching around on the internet found me these highly useful formulae:
Thanks to Andy Ruina of Cornell University for this PDF so I didn't have to format this pile of symbols
The formula at the end is of greatest interest to us, and the symbols are as follows:
  • P = the deflecting force:  what we are hunting for
  • l = the length of the beam in question
  • E = Young's (Elastic) Modulus = ~200GPa for Mild Steel
  • I = Second Moment of Area, details below
There are many formulas for the Second Moment of Area, as it's highly dependent on the cross-section of the beam in question:  we'll be using (Pi*r4)/4 since that corresponds to a solid cylinder.  Our cylinders are 4mm in radius so (3.14*0.0044)/4 = 2.011x10-10 m, there being two rods so double that to 4.021x10-10 m4

With this information gathered, we can deduce that the denominator of this fraction is 48*200,000,000,000*4.021x10-10 = 3860.39 GPa.m4 (Dimensions will be useful later); the numerator will be 49.05*0.5603 = 8.61N.m3, therefore we get a δmax of 0.00223m or 2.23mm down in the centre.  Keep in mind this doesn't take moments into account nor does it account for the bearing spacings.

Wow, this really is looking like 200 steps!  I didn't expect to go back into my university course this soon, but it did highlight the fact that I may need to review my reliance of Reprap-derived mechanisms especially since I don't want my threaded rods to take on an undue amount of mechanical strain.  I'll end it there for today and continue with this lark at some other point, just because I'm tired of entering HTML mode to put in <sup>(Superscript text here)</sup> constantly.  Time for a kip...