3 3
Gideon Yampolsky

Jet Dragon Project

Recommended Posts

I’m publishing the story of Jet Dragon Project – development of jet powered wingsuit system.

It took 3.5 years, over 5000 work hours, and tens of aborted jump attempts, until first successful flight several weeks ago.  During development I consulted many people, including dear fellow late Jarno Cordia, who had previous experience with jet powered wingsuit flight. However the new system was actually designed from scratch.

January 2020

After initial assessment and reading publications on the web, I decided to use two P400-Pro jet turbines from JetCat. Each can product 400N thrust, so total will be 800N. According to calculations it is supposed to be more than enough for level flight, with approximately 25% extra power to allow climbing up.

After receiving turbines several weeks later I assembled simple setup to see how it runs.

First I checked that it responds to throttle commands -

image.png.cb997cae90e8f78c331cda1cf833e937.png

Then run it on the bench -

image.png.ecfea7222ae40fa05a4ae644c76d5d07.png

  • Like 2

Share this post


Link to post
Share on other sites

February 2020

I needed to  confirm that turbines produce identical thrust. If they are, they both can be controlled by common throttle knob. On the other hand, if each turbine produces significantly different thrust, then their controls should be separated, which makes implementation more complicated.

I built mechanical balance system to run two turbines simultaneously and measure thrust difference.

image.png.fd2d4d81cf2bd1e0f44cc7066ba51649.png

image.png.0c9e6019486b170870a035f872347182.png

image.png.b3b2d9fc9e25555e5e9ff15e35641b06.png

image.png.38552b1daf1da85c0278f8e098f6b1ab.png

image.png.6efe35a270f2332b01a66f4f5f88d879.png

Fortunately I found that the thrust of each turbine is almost the same.

 

Share this post


Link to post
Share on other sites
(edited)

 

May 2020

I needed to check that it is possible to carry weight on the foot when flying wingsuit. Combined weight engines, boots, mounting brackets and starter batteries is 13 Kg . My concern was that it will pull legs down and I will stall.

I checked it by series of wingsuit jumps while I'm  gradually increasing weight attached to the foot.

First I attached 4 Kg of weights (packs filled with iron pellets).

Jumped, felt normal. Good. Then attached 8 Kg.

image.png.cde07d6da09807b69bd6ae8a37f8af15.png

With 8 Kg I stalled heavily - it was fall rather than flight. On the next attempt I de-arched in order to keep correct pitch and was able to fly wingsuit normally again. Good. Seems that it would be possible to fly with turbines attached to the feet.

I had find how to attach the turbines . Key concern was about  torsional stress they will apply – their thrust will tend to bend the ankle sideways. Some kind of support was required to avoid it.

First “research” version of how to resolve it –

image.png.1de22ef2e9489e50d6aad8a14e0ce25a.png

I found that tracking shoes did not provide sufficient ankle support.  Snowboard boots were better idea –

image.png.3b4810db4497655a9ef45bdc88ea5a81.png

Which leads to implementation looking like this  

image.thumb.png.bc44ef717143fbd0d9efb60f85ce287e.png

image.thumb.png.57a54d00beb2fe1c742079b6bbf0795e.png

Engine mounting angle is 18 degrees.  In wingsuit flight, legs are stretched apart, this will make thrust vector to be in match with flight direction.

 

Edited by Gideon Yampolsky
  • Like 2

Share this post


Link to post
Share on other sites

September 2020

Next I needed control handle – to start and stop engines, and control their thrust. It has to be ergonomic and easily accessible while in flight. First I built several dummy prototypes from cardboard –

image.png.dc7904fce4be28709da9cb67dff26860.png

image.png.b8ebf0f833cec00cc9e2b74f3215485b.png

image.thumb.png.ee24849a78d1a3d7a3b0fb5f13abfed1.png

And the final version –

image.thumb.png.1cc7fb746eb53b1acd973751b1cf3520.png

image.thumb.png.606ef3dd3a8a3bddc3edda056be081f9.png

image.png.bfaba9ce557692bfcdb931729e8e715c.png

Later, in 2023, I had to completely redesign its concept …

 

  • Like 1

Share this post


Link to post
Share on other sites
(edited)

October 2020

Fuel container. I considered many options of container - on the back, or small containers on sides, etc. Finally I decided that container will be on the torso. I thought that it has to be flexible, to match torso movement. Container was implemented as fuel bladder surrounded by ribbed aluminum enclosure  and nylon fabric.

Fragments of the enclosure -

image.png.8c0afc6b0f05a2a011cbde09e7945494.png

image.png.cea823fb4ea4b43e7d1095c15ebaca40.png

I lined the enclosure with rubber, to protect the bladder.

image.png.dd6c33c6c7c56aade75a399789a26c76.png

image.png.291f715b2414b4760f5d576177692ae5.png

image.png.46ea643f3f71edc2424c93da2694703d.png

image.png.f0661555ccbb78ceda8a103a94b3a0f6.png

First I prepared small bladder, to test welding method -

image.png.ae4155eb6760c0ece067c87114f8e284.png

Actual bladder, 12 liters -

image.thumb.png.983a81d37f7ce05978a5f7d8746d668f.png

image.png.aaeb97f7c0ea211a78b710e5ab728667.png

Preparation of cover fabric -

image.png.ddeba6b5b7d39cd18d9d1b29537d73e4.png

Almost final, first version -

image.thumb.png.9d0799d0d2155e152c1980dd2e9b8d91.png

Design and build of the container was a lot of work.

After container completion I tried to attach it by zippers to the wingsuit, but found that this attachment was not stable enough. I then prepared harness worn on the body and attached the container, through wingsuit, to the harness.

And eventually I have to test it in flight …

image.thumb.png.c4bd81e67fed2d29263a13c05a825341.png

Note that chest strap of the skydiving rig just goes around it, but has no function in keeping it in place. The container is attached to body harness.

I first jumped with container empty. On the next jump I filled it with water to imitate fuel weight.

The first jump with the container was preceded by strong anxiety. I did not know what to expect from that big bump on the chest. But in air it behaved perfectly. It did not add any annoying feeling and did not create any difficulties to flight. Canopy deployment, opening and landing also felt quite normal. Good !

Currently (year 2023) I already have two containers, which both have similar dimensions, but different thickness. The capacity of thin container is  6 litters, and capacity of thick one is 12 liters, which supposed to be sufficient for 8 minutes of flight.

This is newer design of 6L container –

image.png.cfb6fe2f6765cc1080d9fc95016555ad.png

 

Edited by Gideon Yampolsky
  • Like 1

Share this post


Link to post
Share on other sites

February 2021

Humm …   Nobody was ready to take me up !

There are 3 skydiving clubs in Israel. I spoke with all their managers (one of them is my good friend), but it did not work. Israeli airspace is highly congested and heavily regulated. Launching something experimental from the plane requires permits which seemingly impossible to obtain. Clubs did not want to risk their license.

I then tried discuss it with hot balloon operators. Same result.

I considered rising up by unmanned drone. Well, uncoordinated, in Israel, most probably it would end up by being hit by Patriot missile …

Searching for solution was long zig-zagging path. Eventually, only one real option left – paraplane (powered parachute vehicle), with many modifications. So I needed to buy paraplane, learn to fly it, modify it for high altitude and wingsuit launching, find and train pilots who ready to take me up, find how to obtain permits for this.

It took 8 months.

image.thumb.png.53b80274ba8847bb9af94b8330cb2c65.png

Paraplane being built in workshop in Israel. Many special modifications were made in order to make it suitable for climbing to high altitude and safe launching of wingsuit jump. We equipped it with powerful Rotax 914 engine, not typical for such kind of vehicle.

First version of launching pad. This initial version was modified many times until it has become sufficient for safe jump.

image.png.343347a3ab787926437d7eaba910fea6.png

Paraplane being tested by manufacturer -

image.png.bc56715f9ca26b4fd489b3ab94436a43.png

High altitude test flight

image.png.5964d32f9c330ed9bba92548df306a77.png

More advanced version of launch pad being mounted on the vehicle

image.thumb.png.05932676021a52c39b5a0b787ba71746.png

 

  • Like 2

Share this post


Link to post
Share on other sites
(edited)

January 2022

After endless changes, improvements, fixes, trainings of pilots, damages (at certain point paraplane engine failed and we have to replace it - just one of many issues we faced), and discussions with Israeli Aviation Authority, everything was in place for first powered jump.

image.png.d6e3b90dbe2f4cf043b9b24bfbf7ac5e.png

Well, it did not happen …

The system refused to run when in air.  I made about 10 attempts over period of several months. There was always the same pattern – system was fully checked at workshop, then it was ran on the ground right before takeoff. While everything on the ground was perfect, all attempts in air failed. I stood on the launch shelf, gave engines start command, and … nothing happened.

After landing and back in workshop I checked everything and found the reason. One time it was a damaged component in control handle, next time it was malfunctioning cable, next time discovered that I forgot to open fuel valve, next time it was something else. Issues were hiding on the ground, patiently waited until I’m in air and then showed up.

As issues were resolved, new ones popped up. One by one I cleaned them.

Some issues were more destructive:

image.png.6e329d368ca304452e53e5cb196c5a1a.png

Results of one of failed attempts. Engines began start sequence, but due to inaccurate position on the launch shelf jet partially hit the shelf and bounced, damaging the wingsuit.

I cut damaged parts, and using new fabric rebuilt (sewed) it to the original shape. 2 weeks of work. After repair it was looking more or least as original. I jumped to test it without engines, it was flying perfect. Good.  I’m still flying that wingsuit with engines.

I also purchased one more ATC from Squirrel, so now I have backup.

I then changed shelf design to assure that jet exhaust will go cleanly into the air.

Other “interesting” case:

We climbed up, I stood on the shelf, give engines command to start, nothing happened (as usual), attempted several times to restart. Then I returned to passenger seat, checked everything I can, stood again on the shelf, tried to start engines, nothing happened. Then, frustrated, I jumped. I understood that engines will not run, but decided at least to check once again how it flies, down.

But … attempt to start engines took time, this was prolonged attempt, and … or shit … we did not pay attention that we drifted a lot. Only after the jump I discovered that I absolutely have no idea where I am, and even don’t know what is the direction to landing area. I glided in circles, choose open field and landed.

After landing I checked by phone where I’m. 16 km from landing area. In uninhabited place with no roads.  It took 4 hours of complicated rescue mission to pick me up from there.

 

Edited by Gideon Yampolsky
  • Like 1

Share this post


Link to post
Share on other sites

March 2022

Gradually all technical issues were found and resolved. But system did not started in air yet. I reached inevitable conclusion that I cannot resolve it like this. While standing on the launch shelf, I was in need to know exactly what is happening with the engines. Did they receive start command ?  Did they properly responded ?  At what stage they failed and what is the reason ?

Engines are advanced stuff. They have integrated computer which controls them. Computer sends reports in real time. I needed to read, process and display these reports.

So I decided to develop “dashboard” mounted on my helmet, to display what is going on. Engines reports are complicated. It is not for “ok / not ok” LED indicator. You need computer with high resolution display, and you need to write program which communicates with engines and displays information in comprehensive form.

If such thing to implement, it worth to add barometer, GPS, and recording memory, so full flight computer will be implemented.

The need of computer was known from the beginning of the project. I estimated that it will take 12 – 18 months to design it including software. It is a lot of time and huge delay for the project. I therefore hoped that first powered jumps will be made without it, and I will implement it in later stage. But the situation was that without computer I could not make a progress.

Surprisingly it went faster than I originally estimated, “only” 6 months of work …

Prototype tested in lab –

image.thumb.png.6b5c213966620d2e0efe13e0adb34165.png

Computer display installed with the system on the test bench

image.png.c3e5a6b74c8ef8a3a5efae43df6aaefd.png

 

Share this post


Link to post
Share on other sites
(edited)

September 2022

Computer was installed in wingsuit, created its own long series of secondary issues to resolve, but eventually became essential and critical component of Jet Dragon system. It provided reports about engines status and helped to find and resolve issues.

And it helped to clarify, once and for all, that even when everything is okay, engines are not ready to start. They began warmup and acceleration, but could not reach running conditions, and eventually they shut down.

Edited by Gideon Yampolsky

Share this post


Link to post
Share on other sites
(edited)

October 2022

The reason could be too high altitude – low air pressure and cold. Engines start is specified by JetCat up to 9000 feet. I attempted to start them at 11,000.

As well it could be anything else.

I dismounted everything from the wingsuit and prepared dedicated shelf to take the entire system up mounted on the shelf.

image.thumb.png.0cba692d0cc86ff210f640d5a1577233.png

image.thumb.png.20e19c0cd278aea72b5d75a3ecca6dc1.png

I then climbed to 11,000 and gave engines command to start. They did not start. I repeated, reading computer reports. They reach certain state of ignition sequence and then failed. I took the vehicle down to 8000 feet and repeated. Same results. Lowered the vehicle to 3000, and then to 1000 feet. Nothing. So, clearly it is not due to high altitude.

The only remaining explanation was than they unable to start when vertically oriented.  In all ground tests engines always were oriented horizontally.

 

 

Edited by Gideon Yampolsky

Share this post


Link to post
Share on other sites

November 2022

So I built pivoting test bench which enables engines to be operated vertically or horizontally, and enables to change the orientation quickly.

image.thumb.png.01b984a8d9d6bb835708b7d54ee47b75.png

On test bench it became clear that indeed engines cannot start vertically. They are okay to pass initial pre-heating states in that orientation, but after about 20 seconds they must be in horizontal position, otherwise the startup will fail.

Okay, got it. That means I cannot wait on the shelf until engines start, but rather have to see on display that they began the process and then jump immediately.

 

Share this post


Link to post
Share on other sites
(edited)

January 2023

Almost there. The last aborted attempt. That one was expected.

During previous attempts I already understood that the design concept of control handle is not correct. Although during wingsuit flight it was convenient enough, it was nearly impossible to use it while standing on the shelf. Another related problem was that in cold temperature on January at 11,000 feet my fingers froze even if I wore thick glove. I couldn’t operate handle’s buttons.

We climbed up, I stood on the shelf, but I could not operate control handle. Decided to abort. Better safe than sorry.

So I completely redesigned control handle concept. Moved it to left hand and attached to the palm rather than to wingsuit. Then it was possible to wear mitten over it, so fingers would not freeze. I again built several prototypes, until finding optimal solution. I changed buttons and throttle functions, to be able to operate it with just pointer finger inside mitten.

 

Edited by Gideon Yampolsky

Share this post


Link to post
Share on other sites
(edited)

April 2023

Right on the ground problems began. Barometer chip malfunctioned - my dashboard display did not show altitude readings, and no climb / descend rate. Anyway I decided to proceed. I can judge the altitude visually.

I run the system on the ground. Okay. We took off, climbed to 11,000. I stood on the far end of the shelf. Commanded engines to start. They responded properly.

As they reached the state “Accelerate” …  I jumped.

OMG

The initial feeling was like there are two persons working together on common mission. I was busy to stabilize and fly wingsuit. Engines were busy to start themselves. Throwing short glimpses on the display, for my surprise and satisfaction, I noticed that they are making progress toward full operation.

At certain point engines reported full run, 30,000 RPM.  From that point I can control them. I then increased throttle. 50,000 RPM.  70,000 RPM.  95,000 RPM !   Gentle but powerful force on my foot pushed me forward !

I changed body position to fly up. In regular flight wingsuit will climb a little-bit, then will lose speed and stall after a few seconds. But now it was not the case. Despite climbing up I definitely continued to fly forward at significant speed. It worked !

Without altitude readings, I decided to keep larger safety margins, not to fly until the last moment. So I gradually decreased RPM and eventually shut down engines. Changed flight pitch to pre-opening, deployed PC and landed perfectly.

 

Edited by Gideon Yampolsky
  • Like 3

Share this post


Link to post
Share on other sites
1 hour ago, skydiverwannabe said:

This is a great read and some serious dedication and engineering, thank you very much. And here I am struggling on where to start with my chicken house for my garden. Looking forward to read about your future progress. 
 

Thank you for nice feedback.
And don't feel bad about chicken house, I ran into similar situation with my aquarium, and technical capabilities did not help :-)

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

3 3