My DIY-carve

Ima pilot as well but I have to have a big propeller swinging. Only time I’m a glider pilot is when the propeller quits turning. Love what you did

1 Like

Awesome. I fly these also

I went from glow engines to electric. Got tired of the cleanup and hauling big planes to and from the airfield. Now I build with dollartree foam and can just throw it in the back seat of the car

Do you cut out the parts on the cnc? I’m tired so cutting the foam by hand and would like to know how to convert the pdf plans so I can cut them on the cnc. Any help on how to do that would be great


This is one of mine. It would be so much easier to cut on the cnc

1 Like

I agree. If you know how to convert the pdf plans, start a new thread. It would probably help many others

Have built and flown RC airplanes for 25 years but not active since taking my sailplane license back in 2005 :slight_smile:

Just a few of my models, dating 15-20 years back now. Link http://www.halair.com/own.html
My first computer designed model, single EDF MiG29 which I had parts cut by laser (outsourced)
Kitted this one, sold perhaps 150 sets before I sold off the kit rights.

Now I only rely on prop / power to get to altitude, rest is gravity:

Why We Fly:

Fun-Fly:

Not much going on here on my end, really have not had the motivation to get much done as I am burning the candle at both ends - but its alright.

But, today I had a couple of hours for myself and spent it tidying up the CNC. Topic of the day, homing switches :slight_smile:

Have installed homing switch for X/Y, Z will require a little tinkering before it is set.
Since all my previous work have been fully manual, zeroing etc I really haven’t taken advantage of the added functionality of automating pre-carve setup.
I will use them in normally closed (NC) configuration.

Also ran the router power cable through the drag chain plus 2 pairs of speaker signal wires so I hook up the Z-homing switch, LED light and future Z-zero probe.

So during the holiday I hope to have it all set up with homing enabled, soft limits in place and ready to rock into 2018!

Some quick snaps, here showing the X-homing switch CA´d and screwed tight. X/Y/Z-axis will max out before it would impact the switch preventing damage:

Dont drink… Quick fix reservoir containing denatured alchohol for the mister:

Work space, controller mounted on the wall.

2 Likes

Since Christmas is just around the corner, a couple of presents have been arriving from the far east.

My gantry is seriously in need of a rebuild / more solid structure and I am in the process of sourcing parts for this. I have been surprised of what MDF actually made possible but it can never become real good. Instead of laminating the MDF with sheet alu or similar I will be building a 2nd gantry completely with aluminium, 10-12mm throughout. I will continue to use the SBR16 supported rails as funding won’t allow Hiwin yet.

A pre-built Z-assemby will be purchased as it won’t be cost effective to make my own.

The inside width of the gantry is 490mm so a 500mm ball screw was ordered and received.
New toys are fun!

2 Likes

1st of January = quiet day => hobby time.

Wanted to make a top-hat style joystick switch cover for the 4-way tactile switch. It is only 3x3mm and during long flights feel rough on the thumb.
So turned to POM and Fusion360 and drew up my shape and generated a 3 tool job. One main shape for each tool.

Flight stick in question, with the flight computer controls showing in the middle/bottom:

Fusion360 file, three tools and three bodies.
Bottom shape = roughing / planing / alignment holes (3mm 1F bit used)
Middle shape = the 3x3mm cavity for press-fit of the tactile switch column, back side (1mm 1F bit used)
Top shape = Top-side hat. Overall diameter is 11mm (2mm 1F used)

Here the hats are test fitted, close-up at the bottom :


4 Likes

Awesome view. Happy new year!

Same to you @EliasPolitis !
Thank you for your assistance in 2017 :smiley:

Work in progress, making a new control stick to my plane. I am new to organic modelling and struggle a little with F360 so this is good excersise. The electronics shown in earlier post will be housed by a separate part which will be glued to this stick grip. Hence the angled top.

First image show the modelled shape with finger grips plus a border outlined for containing the area of carve.

This image show the raw stock (bamboo) being carved, due to overall depth of cut (21mm) I only had one end mill that could reach so I went conservative on the feed & speeds. 3mm 2F used. First side being carved.

Quick & dirty video of my machine in action:

1 Like

Preparing an upgrade this winter/spring:
1st step, getting the parts into the computer… :wink:

I can swap the ACME-screw out directly with this one but am seriously considering a completely new gantry…
Once that is done I wish to redo the rest of the machine…
To be continued!

1 Like

@EliasPolitis

Revisited the feed rate issue from last year, I purchased a 3mm 1F (Sorotec ProAlu) end mill in order to do aluminium more efficiently. Unfortunately I broke that bit due to miscalculation / over-aggressive cut rates.

After a while I ordered a new bit and have used it for some small work. Feeds and speeds was based on feel / sound.
Revisited the calculation sheet given with the tool :slight_smile: Feed & Speeds SOROTEC PROALU

I recalculated my rates, derated them by approx 20% due to lack of rigidity and came up with :

#1
Bit dia: 3mm
flutes: 1
WoC 1.2mm
DoC: 3mm
feed: 200mm/min
speed: 10000rpm (Makita Speed# 1)
chipload: 0.01818mm/tooth
chip thickness: 0.01781mm/tooth
MRR: 0,720cm^3/min
Air + Mist (Denatured Alcohol)
This worked nicely but some chatter was apparent.

Since chatter is a result of vibration/frequency I figured increasing the RPM would move the cut frequency higher - hoping it would give a smoother run. Then I played with the numbers and came up with this set:

#2
Bit dia: 3mm
flutes: 1
WoC 1.2mm
DoC: 6mm
feed: 400mm/min
speed: 22000rpm (Makita Speed# 4)
chipload: 0.01818mm/tooth
chip thickness: 0.01781mm/tooth
MRR: 2.88cm^3/min
Air only

This worked very well, the test pocket was done dry as my mister made the chips stack/stay within the carve area. Kept dry they would escape during the carve.
Chip thickness was kept equal to the known-good data point in #1.

The pocket floor / wall cut quality on both #1 and #2 were the best I have achieved so far. Happy with that :slight_smile:

Having the bit manufacturer’s data is a very helpful. The most important value is your mm/tooth.

Keep in mind that the cutting speed shown in the second column is the surface speed, not your feed. This is of secondary importance and you can achieve good results with much lower numbers. E.g. in your case, for a 3mm bit to achieve 500m/min surface speed for wrought alu, your spindle would need to spin at 53000rpm !!

It’s impressive how your second recipe which is 4x more aggressive @2.88cm^3/min worked better. The clip below is from a 1/8" (3.175mm vs your 3mm) and 3 flute bit (vs your 1 flute). This is both thicker and stronger (more flutes => thicker core). This is cutting at 1.23cm^3/min, which is about 60% less from what your recipe achieved! I never thought to go over 1cm^3/min with a small tool. Now I’m curious about pushing it further :slight_smile:

https://www.destinytool.com/news/instamachinist-video-destiny-18-viper

One thing I would try to tweak would be increasing the chip thickness from ~0.018mm/tooth to the manufacturer’s 0.025mm/tooth. This also helps minimize the unevenness of chips due to possible runout (i.e. lower runout/bit_dia ratio). And let’s take the MRR to a slightly more conservative (better finish) 2.5cm^3/min.

bit dia: 3mm
flutes: 1
WoC: 0.8mm
DoC: 6mm
feed: 521mm/min
speed: 18426rpm (surface speed: 174m/min)
chipload: 0.0283mm/tooth
chip thickness: 0.025mm/tooth
MRR: 2.5cm^3/min

Yes, that was one number I did struggle with how to “compute” into my head. It was the main msunderstanding in breaking my first high quality bit. Little free spare time prevented serious efforts finding the solution.

Since the main advice with metal work / hobby CNC´s is to use low RPM I disregarded surface speed and recalculated based on actual RPM.

53000 x 1F x 0.025 => 1325mm/min (per spec)
10000 x 1F x 0.025 => 250mm/min
Derated this to 200mm/min since runout will affect chip thickness. I am still running MDF gantry.

The 3mm end mill in question have a 3mm shaft and while I wait for the Elaire 1/8" collett I was using a ebay collet adapter. This will have impact.

Mind you, the #2 was done once, diameter pocket Ø20mm 6mm deep.

I did try my other 3mm go-to bit, a 3mm 3F Helix at 3x Feed but the flutes gummed up quickly, I guess the chips didnt manage to clear fast enough. Bit survided though and is currently soaking in lye :slight_smile:

Bummer, manage to break the bit doing 5mm DoC at 22k RPM @ 400mm/min :(:cry:
No odd sounds or other audible giveaway that I took notice of…so I gather my feed/speeds were a little too high obviously.
With MDF gantry sides there will be flex and some paths/engagement combinations will cause chatter causing the bit to engage to much.
Might be worth mentioning that the bit had a cut edge length = 16mm so it is tall and therefore weaker. It did break at the top.

I have two main factors working against me:
1 - Weak design (decent but not truly rigid)
2 - Cheap ebay collet adapter
3 - Flute length of tool

I have mods in mind for:
1 - Increasing rigidty & weight by adding Epoxy Granite to the MDF gantry side walls
2 - Elaire 1/8" collet adapter purchased, currently in transit
3 - New end mills with shorter flute lengths (6mm) on order, currently in transit.

Wait, 25mm flute length on a 3mm bit can’t be right. This is over 8x the dia.

If your bit is a FSPAL031.0300.11H25 from here, then 11m is the flute length and 25mm is the flute+shoulder length. Since bit dia = shaft dia = 3mm, there should be no shoulder.

The important length to consider is stickout, i.e. the length of the exposed bit from the collet to the tip. This of course is >= the flute length, since you have to clamp above the flutes so as not to damage them.

I thought to ask about your stickout earlier, to run the numbers in gwizard to check for tool deflection (but I didn’t, did I?). Tool deflection increases as the 3rd power of stickout!

For example, in your #2 recipe:

  • if your stickout is 12mm, tool deflection is 0.011mm (or 60% of your chip thickness)
  • if your stickout is 24mm (x2), tool deflection jumps to 0.08mm (~x8) (or 440% of your chip thickness)

The first one is pretty high already. The second one comes in bright red color.

Bad data above, I corrected my post. 16mm flute length.
I am aware of how stickout affect things, but we havent discussed this earlier.

I was using FSPAL031.0300.16 and I would guess (didnt measure) stickout to be 20-23mm or so.

Ordered two FSPAL031.0300.06 (6mm) and will wait til the Elaire collet arrive before continue experimenting / learning :slight_smile: