Page 3 of 5

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Wed Oct 25, 2017 5:25 pm
by SpaceManMat
OverTheTop wrote:
You planning to have cameras setup so you can monitor actual movements?

Not at this stage, but it would not be too difficult to tape a couple of 808 keychain cameras to the airframe. I will see what happens between now and launch.

My though is that video appears to be a very good tool for analyzing performance, especially when things go wrong.

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Thu Nov 16, 2017 6:58 am
by OverTheTop
Airframe Integration
With the completion (sort of) of the main assembly, time to integrate it into some fiberglass airframe.
I had some 4” airframe sitting around so I pressed that into service for this module. It was just about the perfect length to start with!
A bit of measuring and thinking and CADing, then the 3D (FDM) printed drilling guide was slid over the tube. The 35mm holes were drilled using a hole saw, and the pilot holes for the mounting screws drilled. The mounting holes were drilled out to final size and countersunk after the drill guide was removed.


The edges of the holes in the filament-wound fiberglass ended up with some loose filaments of f/g around the edges so they were glued with the usual Loctite 401 and then taken to the final size and dressed lightly with emery cloth.
The airframe was a bit longer than needed and had a test hole I had drilled in earlier that had to be amputated. I dropped it into the lathe and attempted to use my dental drill as a toolpost grinder. Progress was a little slow so I switched to using the cordless angle-grinder and made a nice clean cut.

A length of coupler was inserted in to the correct depth and glued in with the usual West Systems 105/206.
Holes were drilled at the top to accept the NC and at the bottom to mate to the Velociraptor I am using as the test-bed for this project.
A coat of plastic primer and white paint tarts it up a bit. I will add some decals as a roll pattern sometime before launch.

To close out the bottom of the assembly to the parachute compartment from the ejection gases I took to SolidWorks and sketched up a bulkhead incorporating an O-ring, connectors for battery charging, eMatch connections, charge canisters, and breakwire for switching off the VTS when ejection occurs. This was printed in “flexible” material on the SLA printer.
BulkheadCAD1resize.png (235.42 KiB) Viewed 4242 times

BulkheadCAD2.png (124.58 KiB) Viewed 4242 times

Charge canisters were machined up to suit and the wiring connections to extend the interface to the VTS module were fitted.


The bulkhead was fitted to the coupler on the airframe. There was a bit of shrinkage on this 3D printed part (the flexible material seems to suffer from this problem) so the diameter was boosted a little with some Kapton taped wrapped around the boss. Also, the holes were drilled in the fiberglass airframe and the bosses in the bulkhead were drilled, tapped, and fitted with Recoil thread inserts to hold the bulkhead in place


So that is the completed airframe, sans VTS module, and also yet to be fitted with an M6 eyebolt to tie everything to the recovery harness.

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Thu Nov 16, 2017 7:36 am
by SpaceManMat
Looking very nice, I do like the use of that drill guide. Can’t wait to see the final result.

Roll Pattern

Posted: Sun Nov 26, 2017 10:00 am
by OverTheTop
I have added a roll pattern to the airframe section. This will help determine roll attitude if anyone gets sufficiently good pics during operation. Two rectangles of 3M vinyl decal material were cut to size and applied to the airframe. I decided to use a two-bit Gray binary code to encode the rotation in images. Simple and effective.

I thought about higher resolution (more bits) but I think it would be superfluous. Bigger squares are more easily resolved in photographs anyway

I suspect the pattern is purely cosmetic at this stage. I will let you know after the flights if I got any value from it.

The airframe in this pic has been fitted with a dummy servo hub (3D printed) and non-flight fins printed on the SLA printer. This allows this section to show what flight configuration looks like, and still lets me work on the internal module on my bench.

I am now looking at making two (four?) housings to hold keychain cameras to record the fins during flight. Stay tuned...

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Mon Nov 27, 2017 9:12 am
by SpaceManMat
Are those one piece printed fins? It looks like you have a rod up the middle.

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Mon Nov 27, 2017 11:43 am
by OverTheTop
They are SLA printed fins. There is provision for a rod up the center to hold it onto the servo spline and flange, but I will likely not make the required bespoke screws for this. Mechanical strength testing in the next few months will tell me if strength is sufficient. I expect it to be. A screw rod up the center will increase the resonant frequency a little but I don't expect flutter with the design as it stands in relation to the expected flight envelope.

The fins above are not expected to be flight fins.

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Mon Nov 27, 2017 12:14 pm
by SpaceManMat
Snapping a fin on landing may be preferable to damaging the servo.

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Tue Jun 26, 2018 7:12 pm
by OverTheTop
Working Towards Flight One

Make a list, check it twice...
Made up another list of things I need to do to get this into the air. Surprisingly long :(

The original M4 metal spacers holding the fiberglass discs of the assembly together were replaced by titanium bike spokes (custom cut) and some CF strut, ID=2.5mm, outer=4mm square. This saves an enormous amount of mass. Each set of spacers went from about 140g down to less than 7g IIRC.

Provision is made to feed the ejection control (TeleMega) from the NC through the VTS and down to this bulkhead. The NC stays attached to the VTS and this means that the whole assembly is treated like an extended nosecone on the test vehicle. This should make preparation for flight straightforward.

A break-wire on the back of the VTS breaks when the assembly is ejected from the airframe and ensures the VTS is put into a neutral, low current-draw state so the servos are not fried or the batteries drained. This thing can draw around 140W of power if it is provoked! The flight battery for the servos is rated for around 3 minutes of operation at those levels.

A piece of 6mm allthread holds the VTS to the coupler and the bulkhead.

Verification of Function

Now comes the task of working through each of the functions and how it is intended to work, verifying that my expectations are being met.
- Update schematic diagram. Check
- Test that servos actuate fins as expected. Check.
- Verify that there are no integrators winding up in the control system. Check (PD control loops only)
- Check servos actuate as expected in inverted position. SNAFU :(

“Why?” you ask. Remember that I am using a commercial autopilot for this. Thinking about a rocket flight there is an enormous acceleration in the upward direction during boost. I have the AP set to a neutral position during this phase as the accelerometers and gyros cannot be believed. At burnout there is actually a net deceleration on the airframe due to air drag, so the sign of the “gravity field” according to the AP swaps. This is equivalent to our fixed-wing aircraft flying inverted. This autopilot is not particularly programmed for this mode of operation. Any slight inclination will cause the fins to slam into the stops in an effort to flip the bird end-on-end. Not good for my Velociraptor R this way. Data from previous flights indicates about 0.5G in the earthward direction, on top of the standard 1G Earth gravity field.

To get around this I propose to mount the AP essentially upside down in the VTS. This will give it the correct orientation for the coast phase steering, and fins will be in the neutral position for the boost phase as before.

Launch Safety

Part of this journey involves considering what can go wrong and how to keep the flight safe. This will be flying on a K or L, so significant energy is involved. The master plan is to:
- Double the standard distance to the launchpad as it is a complex rocket
- Keep the fins neutral during boost, so the VTS commences active mode when the somewhere over 3000’ altitude.
- Main deployment is set for 2000’
All those things mean that, whatever the result, it happens in the air and the hardware is under chute for the landing.

Next Tasks

So my first priority is to confirm my hypothesis of inverting the AP is acceptable. If that turns out OK then I proceed with integration and hope to fly this in the next month or two, assuming we don’t get scrubbed. Having the AP upside down assaults my sense of what is appropriate, but in the end the engineering definition of being fit for purpose is what matters.

Strength testing of the fins is also high on the priority list now.

Other than that, there are a myriad of minor jobs to get this project soaring.

Stay Tuned!

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Wed Jun 27, 2018 1:41 pm
by OverTheTop
I am actually reconsidering my earlier finding about the behaviour during coast. Need to think about it some more, but too much happening for me to concentrate on that currently. Could go either way still...

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Wed Jun 27, 2018 3:09 pm
by air.command
Very interesting update on this project OTT. Wow, 140W is quite a bit power to have on hand when needed.

Interesting problem with the negative G-s. How will you determine when it is safe to start trusting sensor after burnout?
Also when the servos are in their neutral position, are they actively being held there, or they are just not commanded at that stage?

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Wed Jun 27, 2018 5:06 pm
by OverTheTop
There is a G-switch wired into the circuitry as well as the altimeter control. If the acceleration is over 3G the electronics just commands centering of the servos. I don't want any extraneous movements when the accelerometers or gyros are pegged.

Servos are actually actively centered when the VTS function is not engaged.

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Wed Jun 27, 2018 7:31 pm
by SpaceManMat
You could fly it without fins. Place a camera on the exterior to see how the servos are behaving.

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Wed Jun 27, 2018 8:29 pm
by OverTheTop
You could fly it without fins. Place a camera on the exterior to see how the servos are behaving.

Thought about that, but may as well go the "whole hog" as the pig farmers say :)

Fin Strength Testing

Posted: Sat Jul 14, 2018 5:10 pm
by OverTheTop
Testing Fin Strength
I have access to a couple of 3D printers. One uses FDM (StrataSys 1200es), and the other SLA ( Formlabs 1+) technology. So, the question arises: Which material should I print my flight fins from?

The FDM prints (same as my earlier 3D printed parts) have a weakness in the Z-axis direction due to the method of extruding a melted filament onto the part, layer by layer. I have printed the fins in this machine so the z-axis is parallel to the fin planform. That means it should flex well (full strength lengthwise along the fin) rather than snap weakly where the bending moment exceeds the z-axis strength limit, somewhere near the root of the fin if I had printed it the other way.

The SLA parts are less prone to such weakness, because the resin is bonded by lasers, as each additional layer is bonded to the earlier part completely. This results in a much more homogeneous part.

I really need to test the strength of these parts to see which is the strongest one. That will decide what I fly. I will test the flexibility of each as well.

I had a circular 3D printed part that held fins in the airframe as a display version only. This was perfect for holding the four fins for testing.

The milling machine was set up with a vice to hold the 3D printed parts. Each fin had a pair of holes drilled to accept a stainless-steel wire rope loop that I could hook the force meter (luggage scales!) onto. The other end of the luggage scales was attached to a spring which was then attached to a tool in the quill of the milling machine. So, by translating the table of the mill I was able to apply a controlled force to the fin.

A dial gauge was fitted to measure the tip deflection of the fin, and a temperature meter kept track of the ambient temperature to ensure consistent results.

Here is a pic of Tough as it was loaded up. That is about 22kg sideload.

Here is the remains of the FDM fin. It was a brittle failure, and I still have not found the rest of the parts that were launched somewhere in the workshop.

The Standard Grey failed similarly, as can be seen from the fin stub remaining after it snapped

Here are the first three fins, or what I have of them. You can see there was significant loss of pieces of the Standard Grey due to fragmentation.

The ding in the blue Tough fin was due to it hitting something after the mount let go.

MaterialSLA ToughFDM BlackSLA Std GreyPro Grey
Load @ 1mm deflection1.7kg1.9kg1.9kg2.3kg
Breaking load or Yield loadGood to 40+kg22kg34kgGood to 40+kg
CommentsIntactBrittle failure at fin rootBrittle failure up from root,

Re: Vertical Trajectory System (with some 3D printed parts)

Posted: Sun Jul 15, 2018 4:00 pm
by OverTheTop
Managed to find the broken piece of the FDM so I have updated the family portrait (scale in inches):

You can see from the results table in the previous post that the Tough and the Pro Grey were the strongest of the group. The capability of that material has really surprised me! I may get around to testing them further and pushing them to failure. The Pro Grey is noticeably stiffer than the Tough so is the likely best choice from my available selections. Less chance of fin flutter. I will get a set printed in the next week or so.

Note that the test was performed with a point load near the extremity of the fin. In reality it will be closer to a UDL (uniformly distributed load) with airflow over most of the surface. This means the fin is well and truly capable of what I need. I suspect the servos are the weak link in my control surfaces now.

The current master plan says I will fly this at the next Tripoli launch on 5th August. A K1100 should get it moving nicely on the Velociraptor :) . That give me three weeks to get it completed.