Jetpack Development &Testing

Throughout all my projects I do most of the calculation and design work to produce pencil drawings, for most aspects this is sufficient but where necessary I lean on friend Andy Scott to turn these drawings into 3D CAD. I produce almost all of the hardware in my own workshop - turning, milling, fabrication, welding etc. My wife tells me I spend an unhealthy amount of time in there! Friends Ian & Paul Bennett & Roger Marmion sort most of my electrical/electronic requirements & I receive help from others when needed.

Starting with an almost clean sheet of paper, the propulsive efficiency calculations (mass flow x Velocity v's power required) for VTOL flight will lead you to something resembling a Helicopter. However as a true JetPack is unique in that its possibly the only flying machine strapped to you, rather than you strapped into it, and is expected to have no visible wings, (fixed or rotary) the rotor & pack has to be very compact. This high power to density ratio comes with the unfortunate trade off in efficiency.

Although I believe sorting the engine represents about 75% of the challenge I have spent considerable time also studying the flight dynamics and control issues. However having satisfied the fundamental design requirements it becomes a bit of an "egg before chicken" scenario, you can't test the control systems until you have a safe functioning lift engine with the right layout and dynamics. Initial testing of the flight controls will be performed using a purpose built test rig, tethered flight will then follow once the system has been perfected on the rig.

I decided very early in the design that I was prepared to manufacture everything except the turbine rotors, I also wanted to avoid scratch building the compressor, fan rotors and combustor. These parts, particularly the turbines are highly stressed components. To produce rotors with the required level of performance and integrity & subject them to the 'type testing' normally associated with man rated flight would be prohibitively expensive and beyond my budget. I therefore chose to source these parts from existing, preferably man-rated engines that could be secured as military surplus. However identifying & then matching suitable parts from different engines & conceiving a design to bring them all together proved very difficult, just this part of the design process took 4-5 months! Having finally settled on a design, the chosen engines containing the rotors had to be found, this entailed travelling around the UK and Europe, taking them back to my workshop to remove the said parts, evaluating and modify the items not rejected to suit my design, (obviously modifications to rotors had to avoid any stress raising) and often constructing elaborate test rigs for further evaluation. I was so busy doing all this I am now horrified at how few pictures I have, those used here & some data records are sometimes the only archive of this work!

MK1 (stood on a roll of welding wire) with me dreaming of flight!

The first (Mk1) design was actually the most complex of all those I conceived - a twin spool, co-axial, contra-rotating, low/medium bypass turbo fan engine with 7.5:1 compressor ratio, two stage upward facing fan supercharging a Titanium centrifugal compressor, with engine driven accessory gearbox driving scavenged/recirculated oil system & fuel pump, detachable electric start etc. It was also exceptionally compact.

I spent almost two years designing, building and testing this power plant during which time I sorted out several mechanical, thermodynamic & aerodynamic difficulties, and did run it to 100% several times. I effectively scratch built about 65% of the unit. However it had a basic problem - the "swallowing capacity" of the HP turbine was about 15% -20% less than I had calculated. This had a compounding effect, with the HP spool at 100% the LP spool would only reach 70%. I tried to look for a replacement turbine but nothing was suitable. I had originally chosen the turbine because it was intended for co-axial use, i.e. a large hole in its centre allowing my fan shaft to pass through, so knowing it was the only possible candidate in this configuration and after a great deal of reluctance I had to admit that MK1 was not going to provide the necessary lift. Still I learned a lot & so had improved my ability for MK2!


Two stage fan assembly from MK1 being tested. I used a RR Gnome helicopter engine to drive it.

After considerable study and the use of several purpose built test rigs I decided that due to many technical benefits my future designs would use a larger more efficient single stage fan mounted just above & rear of the pilots head, this meant I had to generate less power (still 325HP!) to drive it but tolerate a bulkier yet acceptable package.


Ready for yet another test, this time using a Noel Penny 100 series gas turbine to drive just the first stage from my two stage fan. This allowed me to alter stator vane angles, duct designs, flow straighteners etc. Throughout this project I have built several different test rigs, some almost become a project in their own right but verifying certain design calculations is essential.

Mk2 was partly intended as a viable power plant but also as a test bed for the new Fan and associated hardware. A somewhat simpler system, it had the single stage fan facing upwards supercharging a single stage gas generator, in fact a substantially modified Microturbo TRS-18 turbojet. This turbofan was again run several times to high speed with only a few mechanical teething problems which I sorted out. All of these tests (pictured below) were performed in the horizontal position.


Mk2 ready to test, the first view would be looking at the bottom of the pack, and this view from the side.

However after about 1 year's effort this configuration had three ultimate problems. The main problem was that the LP turbine was not matched. With the fan at 80% the gas gen was at a perfect 83%, but the EGT had reached max. The second problem was my dislike with using the TRS-18. Although this is a superb small light engine it was intended as a limited life drone unit. If I slightly derated the engine to make it more suitable for this manned application it had hardly enough power. Modifying the oil system to run vertically was certainly not going to be straight forward and with spare parts not available unless you are the MOD I was beginning to doubt it's suitability. The third aspect was that I simply did not like the awkward layout with the hot & cold exhaust both exiting at shoulder height but due to pressure differential & temperature issues they had to do so in separate ducts, this and the ductwork required to supercharge the gas gen made it a cumbersome layout. So again I had learned a lot but this was not the solution, it did however serve it's other purpose - a good initial test bed for the larger single stage fan and its associated assemblies for use in MK3.

I have been working on Mk3 since early/mid 2005. (pictures will follow at a later date) As with MK1&2, the basic design weight & thrust is:

Complete dry pack = Max 150lbs
Max fuel capacity = 50lbs
Max thrust (at ambient 15 C) = 450lbs
Max pilot weight = 210Lbs

With the reasonable fuel efficiency measured from my turbofan, the design goal of 10 minutes flight duration is reality!

MK3 is on target to achieve these design points, and has been run numerous times to 100% I have sorted out most of the teething problems and I'm continuing to refine it. Throughout this continuous testing I am of course also 'type testing' it. I've measured large losses in the thrust ducts which I am working on, despite these it still makes the power but does so at a slightly higher than ideal EGT. Improvements to the system will allow me to back off the engine from max to a more acceptable 95%, whilst still producing sufficient lift. All of the flight control mechanisms are designed to an outline stage. I will upload some pictures & possibly video at a later date, I hope to 'finish' the engine and control systems throughout 2007 and all being well begin rig & tethered flight in 2008. Watch this space!

22/4/07- Update. In the few months since I wrote the text above (when this site went live) I have been working long hours on the engine/pack and have completed several good test sessions. I have modified the design and totally re-made the fan stator assembly & fan shroud to a "finished stage" with excellent test results, including a valuable weight reduction of 6LBS. This assembly was previously a combination of parts taken from the fan donor engine & parts I made to adapt it for testing, knowing that a lighter more compact finished assembly would be required. I have also finished machining the accessory gearbox which will drive the oil & fuel pumps, these pumps have previously been driven by electric motors to allow sufficient data to be gained from testing before attempting to design and make the finished versions. I have also added & tested many more minor details such as a FOD guard for the gas gen, custom built oil tank, tweaks to the oil system etc.

During all of this engine work a CFD analysis is also being undertaken to look at the flow patterns within the thrust ducts to try and narrow down the cause of the losses I have reported above. This analysis continues and I will provide results when complete. I will also be adding some pictures of the engine/pack and some video of my previous projects in the next coming months.

10/7/07 - Update. Good progress continues, the accessory gearbox driving the two pumps is now fitted and a slightly revised ductwork that delivers the fan air to supercharge the engine is also complete. Both of these in their "finished" state have been tested. I am just completing fabrication of a revised Titanium bifurcated exhaust for the hot exhaust outlet. This section is fairly complex as it also supports the bottom main bearing, many other smaller parts have also been completed. The CFD work on the cold fan air ducts is hopefully nearing completion, its being done as a "Favour" by a friend who works for an F1 team, this has meant him doing a little when the chance arises so I have to remain patient. I am keeping these updates brief at this stage as much of this is probably difficult to follow without pictures, these will come with the next update, so stay tuned!

8/10/07 - Update. Continued testing and development over the past months has produced a near complete "flight ready" engine! - see the pictures below.

Although the results from CFD work to study the duct losses were never completed a few indications of possible loss area's were gained, this info combined with further practical tests using the sheet metal ducts allowed me to revise the duct design reducing the flow losses to a satisfactory level. I was then able to produce the glass fibre ducts seen above, a carbon fiber set will be made once all flight controls have been proved. Obviously these pictures show the unit at its current stage prior to manufacture of the body corset or flight controls, the left picture is viewed from the side the pilot will operate it. The body corset with a multipoint harness will be similar to that used by the rocket belts, (see history section) with the pilots head located in the inverted 'V' created by the lift ducts, the front outlets protrude either side of his shoulders. (see picture also added on home page)

The hot exhaust exits from the bottom of the engine and is diverted either side of the pilots legs. Cold air from the small rear fan outlets (visible in the middle & right picture) then mixes with the hot to reduce the mean temperature. Almost all of the unit has been custom designed & built by me for this one project with only the turbine rotors, part of the combustor, part of the compressor and fan rotor originating from several different other man rated engines. This very difficult development has now taken the project to 5 years of constant effort but I am at the exciting stage of beginning to add and test the flight controls so keep watching for progress reports!

22/12/07 - Update Progress continues as planned, as seen from the updated picture on the home page initial flight controls have been engineered and added, I have also begun testing these for control power using my static test rig & digital load cells, so far the numbers are spot on! As with most aspects of this project because I've started with a "clean sheet of paper" everything has to be worked out from first principles, I have obviously gained much valuable information from the Bell work and other technical papers but with the layout I have chosen I have a unique Jetpack. All aspects regarding the body corset & harness will be refined, however the existing fiberglass corset is also part of the next test rig that will be used to refine the flight controls before tethered flight. Testing so far has been done without on-board fuel tanks, this will also continue until I begin tethered flight for reasons of safety and convenience. With assistance from friends we have also moved much closer to a "flight ready" engine speed governor & data logger set up, as always testing continues.

22/2/08 - Update Performance & mechanical development of my turbofan engine is now complete and ready to start flight testing, however the electronic speed governor, data logger & alarm system being developed by friends Ian Bennett and Roger Marmion is not quite "flight ready", we hope to have completed this within a few weeks. When these electronics are finished and tested, this will conclude the "static test program". On the right is a recent photo of the static rig just prior to one of the final engine tests. This simple rig has enabled thrust (lift) & torque (yaw) to be measured through digital load cells along with all the other engine parameters such as pressures, efflux velocities, flow rates etc.

The next phase is to dynamically begin refining & testing the flight controls. I have built a multi-axis test rig for this purpose that will enable the man/machine combination to fly in hover with freedom of movement through the relative CofG in pitch, roll, yaw & vertical axis but within the constraints of the test rig, this should greatly reduce the risk of complete loss of control or injury during the initial flight testing, this multi-axis rig should be equivalent to a moving flight simulator (in hover) & maybe better as it uses the actual jetpack.

8/5/08 - Update With a project as complexed as this there will always be some hold ups, the last couple of frustrating months has produced little real progress. I hoped to be reporting on my initial flight trials using the multiaxis rig but the electronics require further improvement before we can progress to the flight program, however progress viewed over the past 18 months has been very good and I hope to report more next time.

3/9/08 - Update Flight testing has now started ! also see the new pictures on the home page! The initial tests appear very encouraging, particularly following the difficult period reported above that delayed progress by several months. The multiaxis rig connects to the jetpack at the vertical & horizontal CofG, the rig then allows some movement in all axis but each axis can be locked out to allow the pilot to learn control of pitch, ascent, decent, roll & yaw one at a time if required. The thrust vector, control power and man machine dynamics can also be more easily studied. The rig has been kept light weight to reduce it's inertia & is neutrally balanced, it is essentially a restraining device much like a tether and will be used for several months during the early flight tests. As testing with the mutiaxis rig is performed on the same spot, after a short period hot air recirculation and upwash buffeting are potential problems, so I have fabricated the raised grill platform to help allow the efflux to further dissipate. The complete pack is spot on its design weight allowing me to walk short distances with it strapped on and switched off, however with the pack attached to the start trolley the first operation before testing or flight is to start the engine and allow it to idle, it then more than offsets its own weight before disconnecting from the start trolley.

Update 21/1/09. Following the first series of flight tests (reported above) an opportunity to get the thrust ducts re-made in carbon fibre through a friend working for an F1 team could not be missed, particularly as I'd also concluded that a small modification was required to balance thrust lines relative to the CofG. This meant using the glass fibre ducts seen above with modification as the patterns for the new carbon set, so this meant a hold on further testing. Unfortunately the ducts were not completed, although much helpful work was still done. However a new revised set of glass fibre ducts are now completed and during this 'down time' I have made several small but worthwhile improvements to the gas generator (engine) resulting in further weight reduction and slightly improved performance.

The winter weather can make testing a little difficult (muddy access to the test site etc) but despite this I hope to continue flight testing in February, and I expect to achieve free flight during this year!

20/08/09 - Update Flight testing within the multiaxis rig continues. As seen from the video above, control in pitch, yaw and throttle progressed well and further improvement has continued. I designed the "first cut" version of flight controls as the simplest to engineer as I felt this was the best starting point, I also expected to-much control power in yaw and to-little in roll. The multiaxis rig allows me to individually lock out yaw and roll, and since the video was added the last axis- roll has been unlocked, as with yaw and pitch the controls required modification, roll has actually proved the most difficult to get good authority over but after many small mods & tweaks I am now getting there. It is worth noting that not only am I having to learn how to control it I am simultaneously having to develop the flight controls to enable me to control it!

Further testing/practice in the rig will continue until I am confident that I have sufficient control to allow a safe progression on to tether. All of the tether equipment is ready, with the need to eliminate any chance of a tether line being ingested by the fan this aspect also required much consideration.

Jetpack Development & Testing