Homebuild Cruise missiles

reply to post by rat256


Rat - you made one of those too!? Awesome! I have two of my own, its why my avatar says "DIY jet engine guy".

Planeman - have you thought about using a Wankel rotary engine? I own a 2005 Mazda RX-8, it's the only production car in the world to use one of these unique engines. And man, does it go.

www.youtube.com...

Above is a youtube link explaining how a rotary works. Rotary engines have a lot of advantages, they are smaller and have less vibration, but the biggest advantage is that their torque builds as their RPM builds. It doesn't drop off like piston engines...if you drive it hard it just begs to be driven harder. For instance, my RX-8 will scream all the way up to 9500RPM if it has to and won't start dropping off until I'm well over 9,000.

The downside is that it has low torque at low speed and it really drinks fuel and oil.

One of these engines might be useful in a small missile driving a prop or compressing air for a turbine.

[edit on 17-3-2008 by BlackWidow23]

reply to post by BlackWidow23


RX-8 engine isn't that special performance wise. For example BMW M3s new V8 goes to 8500rpm... but how heavy is the wankel engine? Compared to lets say 1000cc motorcycle or 4-cyl car turbo engine. For a missile use the critical factors are weight and fuel consumption.

If we having a range of 500Km or more it will have to be air launched or ship launched, as there is a treaty (NATO and former soviet union countries) banning ground launched cruise missiles with a range of 500-5500 km

I think we probably need a thrust to weight of about 0.25 for the missile, as that is about what the Tomahawk has and also 2 engine airliners when on 1 engine.

Take the AMT Titan jet as posted by planeman

The AMT Titan UAV has a thrust of 392N converting that to pounds is 88lbs, that give us a weight for the cruise missle of 352lbs.
Take of the weight of the AMT Titan 4.575Kg (10lbs) leaves us with 342lbs.
Now add a SDB at 285lbs (give us a warhead, guidence and wings) add a light weight plastic pressurised fuel tank which leaves us 57lbs for fuel.
The AMT Titan uses 1.12Kg per minute so that gives 23 minutes powered flight. If it can manage 800km/hr like the Tomahawk, that's a range of 308Km. The SDB has glide range of more than 60nm which is 111km so I guess we can add that on which gives us 419km, probably less due to the extra drag. I guess we need more thrust so we can carry more fuel, You might be able to get more thrust out of the titan at the expensive of it's operating life and we don't need it to last long.

Turning a jet into a turboprop seems to give you about 4 times the thrust for about the same fuel usage, although I haven't seen one based on the bigger model aircraft jets, also it'll mean a slower cruise speed.

I'm over budget as well.

reply to post by BlackWidow23


I did computer systems engineering at uni and as part of the first 2 years we were lectured by a chap who developed jets for rolls royce to give us a general feel for the world of engineering and as part of his lecture series we got told that the DIY jet engine was possible and that we had the whole of that term to make one... i managed to get mine to Mk3 based on a huge garrett turbo from a truck. Mk1 was a small turbo from a 1.9litre diesel engine it worked OK but i kept burning out the waste gate that i was using as an overrun protection. Mk2 was a pair of small garrett turbos from a nissan 300 with one shared very large combustor (a 25 litre oil drum) Mk3 was a turbo from a mack truck using a pair of the biggest sparkplugs i could find and a comercial baked bean can, with a smaller comercial fruit-in-syrup tin as the perforated interior as the combustor!

its all good fun! i have a plan to make something that flies using the Mk2 twin turbine engine as it provides an obscene amount of thrust and seemed quite reliable!

reply to post by northwolf


What everyone drools over in a wankel IS its size and weight, and the fact that it pumps out massive amounts of juice for its size.

The average 1.3L piston engine might pump out 120HP at absolute best. The 1.3L wankel engine on the RX-8 pumps out 232HP.
If we were working with a piston engine we would need a fabulous 2.0L I4 or a decent V6 to get the same power.

My point is that despite its thirst, the wankel packs a helluva punch for its size. It might be good in a DIY cruise missile.


Rat - good to know of another DIY turbocharger engine maker out there. I don't think I'll ever get airborne but I am working on a gokart powered by a propane engine, medium turbo with 3.5" wheels. I guess I might get airborne someday but I will need to finish school and get a degree in aircraft engineering before I attempt to create a stable and controllable airframe.

reply to post by BlackWidow23


Hey if your building the engines anyway why don't you look at WIG craft? (wing in ground or Ekranoplan's)



Got to be easier to build than an airframe intended to get real height (your only really looking about about 10 meters max with small WIG's) that would be a cool thing to do with a home made jet

reply to post by Now_Then


ooh, double post my bad. read below!!

[edit on 18-3-2008 by BlackWidow23]

reply to post by Now_Then



Haha, wouldn't THAT be nice. Unfortunately my best engine only pushes out about 90lbs of thrust and weighs about three times that. Needless to say I doubt I could ever get a T/W of over .2 or so if I was lucky which really isn't enough for much of anything. That, plus its pretty huge, about five feet long and two feet tall, with the turbo itself weighing in at 100lbs.

And thats without fuel.

Regarding the building of a home-grown cruise missile,
engines aren't too much of a problem, since any Ski-Doo or Sea-Doo
Rotax-style would do as the basis of a ducted fan engine.

Most competent machinists (ME!) could build a high-performance
fan blade system just by scanning online photos of any Pratt & Whitney
or Rolls Royce 747 or 777 engine and then scaling DOWN the fan-blade
design to the dimensions needed to fit in my tube.

I must admit the ShowStopper for most people would be be avionics
and Flight control. Most people do not have the expertise to create
real-time software to do live flight control, avionics & fuel management.

BUT -- Just to brag more than a little....in the last 10 years I have become
VERY EXPERT in creating NOT only real-time flight control/avionics software,
I've also become EXPERT in creating completely autonomous
GPS and Vision recognition software that can do Ground-Hugging
and Terrain Following flight.

In fact to BRAG even more, I've got a cheap RC (hmmm maybe
NOT that cheap!) helicopter that uses my very own designed & coded
Object Pascal source code that is currently able to follow rivers, roadways,
trails and even IDENTIFY, TRACK and then follow MOVING people
and Cars UTTERLY autonomously using Multi-Spectral Machine Vision
WITHOUT needing the use of GPS coordinates.

I can just take a Google Maps Satellite photo, draw a route on it
with a felt-tip pen and my craft will follow the trajectory outlined and
can AVOID birds, powerlines, buildings and terrain using just the
machine vision & object recognition technologies I have programmed.
NO GPS NEEDED WHATSOEVER!!!!!

And I can do this at a full 24 bit RGB colour
at 1920 by 1080 pixels at 60 frames per second
progressive scan at a maximum of 160 miles per hour
which is DAMN GOOD for a home-build craft!!!!

I'm even learning how to home-build multi-scan and SAR radar
so I can up the flight speeds to supersonic rates for autononous
navigation and flight control and automated collision avoidance
of very high speed home-built jets.

ON A WAY TECHNICAL NOTE:

The higher the flight speed you want, the higher the computer horsepower
you need to do machine vision and terrain following for flight control.

My BEST system uses 16 processors I dug out of six Sony Playstation-3's
I bought so I could get the IBM/Sony Cell processors which have
PHENOMENAL DSP horsepower and I then designed & built a custom
motherboard with 64 Gigabytes of RAM and my own custom built
multi-processing BIOS plus my machine vision and my real-time
Linux-based avionics software.

Most general purpose CPU's have a hardware interrupt latency of
only 16 milliseconds which lets me do 60 frames per seconds or
about 160 MPH maximum for real-time terrain following flight.

I've been able to do Microsecond interupt times on the CELL processor
(WHICH IS NOT SUPPOSED TO BE PHYSICALLY POSSIBLE from what
I've been told by IBM) so I can get 1000+ frames per second which will allow
me to do over 2000 MPH for autonomous terrain following flight
using only vision recognition in 1920 by 1080 incoming video feeds
at 1000 frames per second!!!!!!

Thus it is VERY POSSIBLE for a civilian to create MILITARY QUALITY
avionics and flight control....BUT.... you have to INVEST the time and
money into it and at $90,000 and 10 years & counting, in addition
to some painful failures it's been a damn expensive and frustrating
hobby...BUT IT CAN BE DONE !!!!!!

So YES I could build a home-built cruise missile that would in fact
WAY OUTPERFORM a million-dollar Tomahawk....but I have MUCH
better USES for my time time and energy...like building autonomous
digital mapping UAV's and open-source mini satellite systems.

With time, perserverence & skill and reasonably coompetent engineer
COULD DUPLICATE what I've done simply by modifying source code
already out there for flight control software, engine designs and
airframe hardware to create anything from cruise missiles to
personal jets.

Regarding my, previous post, I was in a bit of a "Rotten Mood" yesterday
due to my allergies, so those comments sounded rather "Braggy"
and Arrogant - so my apologies for that!

But I did want to add some technical clarifications....

1) MOST general purpose CPU's such as AMD/Intel/Via use something
called hardware interrupts in order to do multi-tasking on a
Time-Critical basis which means that a processor can within a
guaranteed 16 to 32 milliseconds time-frame switch over to a
new task, execute some instructions and then go back to the ORIGINAL
tasks.

Because Most Pentium/AMD Athlon CPU's have up to 16 hardware based
threads, this means 16 possible separate tasks can be run at the same
time and each task is given it own memory and CPU slices without
interfering with each other -- This would allows me to play a video,
download a file, write a book and browse the web without those tasks
causing crashes, slowdowns or harm to each other.

3) The problem with the AMD/Intel CPU's is that 16 to 32 milliseconds
interrupt time-slicing for thread execution is NOT a fast enough time
period for me to do mission-critical tasks such as flight control or avionics
or vision recognition at high video frame rates.

4) What I have therefore done is use the multi-core architecture
of the IBM/Sony Cell Processor which I've disembowelled from
cheaply purchased dead Playstation-3's (Don't Worry - I tested all the
CPU's themselves thoroughly!) to create a single threaded
application architecture that is synchronized among multiple CPU's so
that incoming 1920 by 1080 pixel, 24 bit colour images are broken
down into sections that are "Gridded Out" to multiple CPU's which
can then use my multi-processing techniques that do
edge detecting using Integer and fixed point based convolution
filters, then object and vision recognized using template-based boolean
logic to find terrain feaures such as rivers, roads, houses, apartments,
powerlines, trees, people, cars, trucks etc in real-time.

Because I am using 16 CPU's along with the multi-core internal execution
units of each Cell processor, I can get absolutely FANTASTIC
digital signal processing performance at the rate of 1000 fully
edge-detected and Object Recognized frames per second
at full HDTV resolutions using a single forward-looking
high frame-rate HDTV camera.

The BANDWIDTH to do that is 6,220,800,000 bytes per second
or 6.2 Gigabytes per second so that is why I break the task down to
16 processors which at 388 megabytes per second per processor
is easily handled by the individual Cell processors per frame buffer.

Because I need 3 full frames as a edge-detect and motion detection
buffer and 1 frame for an object database template buffer, my total
bandwidth is over 25 Gigabytes per second BUT even still, it is STILL
within the performance abilities of 16 IBM/Sony Cell processors
working as one big supercomputer.

You should see the small turbine generator I had to build just to POWER
the Cell-based motherboard while in flight....!!!!!!

5) And for thsoe technoids out there, I've been able to showhorn the
thing into a kit that is basically the size of a 1/5th scale Bell Jet Ranger III
which makes it is a VERY LARGE autonomous chopper.

I'm using Rotax engines and they're NOT light so this is no
cheap 60-Series chopper.

Back to my intentions, I'm attempting to build BOTH a standard Jet-Ranger
type camera chopper AND a circular Avrocar-style Ducted Fan-based
UFO-like UAV camera platform that will EVENTUALLY be able
to goto heights above 80,000 feet and use VERY high end
custom lenses to take gyroscopically-stabilized, high resolution
22 megapixel still photos from 80,000 feet. (Hasselblad Digital Back)

I'm not quite where I want to be yet....BUT....

my flight control software DOES work magnificently at those
1000 FPS frame rates, so I am confident that a fully
autonomous ground hugging and high-flying UAV can be built
on a civilian hobby budget of $100,000 !!!