I wanted to build a flame tank case mod which was able to drive and off-course throw some nice flames. (to test your insurance policy )
Previously I created another flame throwing robot from the older C&C tiberium sun genre, Both tanks will be using the same custom made electronics.
But now I wanted to make a mod from the new C&C genre (C&C tiberium wars)
I chose the new Advanced flame tank because I really like its design.
The design is kind of retro to the tiberium dawn flame tank.
The old Tiberium dawn flame tank:
The newer tiberium wars (advanced) flame tank:
The Design
I wrote down some cirterea I (at least) wanted to implement n the tank:
An (older) FAT PS3 running Ubuntu linux server distro in dual boot with XMB (when not using the tank function but playing GTA or so
Tank has to be able to play blue ray.
Tank has to be able to drive autonomously(using camera's and advanced object detection techniques.
It has to be completely wireless, no power cables or so.
A docking station is needed to recharge the batteries(like the ROOMBA) and to automatically connect to the HDMI cable and the beamer
2 real working self igniting flame throwers.
The body of the tank at least has to have 180 degrees rotation freedom.
I want to shape everything out of aluminium and Stainless steel.
All electronics and software is designed by me.
These are rather high demands and it probably will take some time in creating and fabricating all components of the tank.
But that is half the fun
I started by drawing the tank in inventor, before this project I never worked with this program before, but along the way I learned how to use it.
I designed the tank from very few designs (couln't find any real good high quality images).
This has some pro- and cons:
A con is, I wouldn't get exactly like what the designer at the time had in mind.
A pro is that it leaves more open to my own creative input...
These are a few screenshots of the in-game tank and a art disign of the tank I found on the internet.
This Time I wanted to make my own tracks and I tried to recreate it as much as possible by looking at the design.
I immediately disigned the matching gear for the tracks along the way.
The tracks will be casted out of aluminium, and in This drawing you can see 12 track connected to each other with special channels.
After this I started on the tracks themselves, I wanted shock absorbers and a chain tightener just like a real tank has.
In total the tank will get 4 of these tracks.
With both plates on:
And te covering plate:
These are my own design shock absorberd which will be used in the tank, most parts will be created on the lathe.
Each track will get 8 of these.
The bottom part of the tank will look like this, the tracks are connected to each other with this frame.
Each track has at least 1 degree of freedom to rotate to allow for maximal terrain accesibility.
An naturally the body itself...
I mainly created these designs to measure from .
And this is how It will get to look.
My own driving PS3 flame tank (who else can say that ? )
In total the tank will be about 70 cm long, 50 wide and 26 high.
I hope you like this thread as I am going to use some nice and advanced techniques to get this monster on the road
Because I designed the tank in CAD I now can just simply export various parts.
I know a guy who has access to a really neat laser cutting device and he agreed to do some work for me.
The results are really crisp clear.
I plan on making a few aluminium tank components myself.
Among them are a number of difficult shapes like:
A ball bearing mount
The tracks
Cover plates
The idea is to make molds out of wood and then make aluminium casts of their shapes.
I have done some research, what is nessecairy to melt aluminium and how to cast it.
Aluminium melts at a temperature of 660 degrees celcius.
And with a normal gasflame it is difficult to obtain temperatures like that (or higher).
That is why a furnace is needed.
I shopped for some special materials like insulation cloth which withstands temperatures up to 1400 degrees.
Special fire resistant cement.
A crucible to melt the aluminium in.
Aluminium cleaning material.
At a local hardware store I bought a steel bucket to build an furnace.
In the top of the bucket I drilled a hole to let the gasses pass through.
The bolts will be used to hang the bucket in a frame.
The frame is made out of some old scrap metal.
The completed frame...
A note to myself, this is why you wear protection glasses.
I made some sort of suspension bridge construction to lift the bucket from the crucible.
This is how the bucket goed up, it can be attached to the frame so you have your hands free while casting.
This is the special insulation which withstands high temperatures.
I've drawn the inside of the bucket on the the insulation.
After cutting it out, some cement on the cloth and the inside of the bucket.
Then I stuck it to the wall of the bucket and had it cure for a day.
Meanwhile I had begun to make the tools to handle the extremely hot crucible.
I bent some rebar and welded some plates to it to make an axle.
Welded it..
A special bracket is made to attach to the plier, this will prevent the crucible from falling out when casting.
This is how the plier will sit on the crucible.
Thats all for this update.
Next update is of the refining of the aluminium.
I am really curious about the results
Meanwhile I asked a friend if he could mill the tracks out of wood, to use as a positive mold in the future...
Update 3 - refining aluminium and first sandcast pour
I wanted to refine my aluminium first to obtain a reasonable quality by mixing good quality together with scrap aluminium.
This is the setup of the system, 6 bricks are used to keep the funace extremely stable.
and 3 heat resistant bricks are used to set the crucible on.
The torch is now on, the best thing is my IR -thermometer indicates the outside wall doesn't get very hot, this insulation is doing its work.
You can see the hot flames rushing out of the pot, things are getting hot now!.
I don't know what this black stond exactly is, at the shop I was told it is used to clear the aluminium of any oxide.
When sturred through the aluminium mixture the inpurities will start floating at the top.
This should be scooped off , till you see nice shiny liquid aluminium.
Now I am wrapping the cleaning agent in the aluminum foil so it can be dropped in the crucible.
When working with liquid aluminium: DON'T forget to use you anti brain scanner hat.
Next I mix and stir the liquid aluminium a bit so the impurities start floating.
Then the impurities can be scooped from the liquid.
Then the aluminium can be cast into forms I use : "mom's patented cake cups "
The aluminium starts out shiny but a dull film starts to develop very fast while it cools.
a nice video of the whole proces.
You can see the crucible is kind of hot.
After some pours we have nice refined nuggets which can be used later...
sand casting and first pour
I want to make sandcasts and thus I needed some tools to create these casts.
I started by creating some casting molds made from plywood, the casting molds excist out of 2 parts which can be lifted from each other to remove the molded object(s).
The mold is marked so you wont make a mistake, how it is put on.
Next up were 2 small (conical) wooden objects which are easely to remove from the cast.
one is the pour and the other is used as riser.
Now it is time for the first cast, this gear will be casted and will now function als the mold.
I dispense some graphite on the mold so it can be removed easely from the sandcast.
I use this red (oil containing) sand the create the sand mold.
Next the sand needs to be hammered onto the mold till is it rock solid and no more bubbles exist.
This is now rock solid but the rest has to be filled to prevent the sand from falling out when ramming the other side.
The excess had to be removed to leave a flat surface.
like this...
Next the mold needs to be flipped over.
next the top has to be filled, you can see why the riser and pour are used... they are sitting cold on the gear here...
After ramming the other side the casting can be removed again ... now the remove the mold itself.
You can see the riser and pour on this photo.
They will be removed before the pour...
Heating up the furnace...
This is the tricky bit.. removing the mold itself... it can be a bit sticking to the sand....
Now both parts are put back on each other... the can see the graphite and some shape of the gear...
The aluminium is hot now ... so lets get ready....
I have made a movie of another pour... but is it very explanitory...
You can see the sand smoking...
The gear can be seen now.
The black sand has been burned and cannot be used anymore ,,, the rest is still good.
You can see the result here... not quite as sharp as the original ... but you can make out they belong to each other
Hello people, finally update number 4.
Sorry for keeping you waiting, my fathers engine of his boat broke down and I had to overhaul it.
And after it was fixed it ran well for a couple of hours but then one of the injectors broke down.
But to the point, I have also done lots of work to the tank.
I stripped my PS3 from its housing to take a look at the electronics
The electronics and pneumatics had been bought.
a succesfull pour of a row tracks have been done.
a track has been assembled.
multiple parts of electronics have been designed
I started shaping pieces of metal.
This is 1 of the airbottles I will be using, it is resistant up to 200 bar pressure and used by paintball devices.
A pressure reducer will be connected to this bottle.
A converter to connect the hoses to the bottle has been bought.
To connect the bottles to the frame I created 2 mounting brackets.
As you can see it is placed on the bottle like this, it is held in place by 3 screws.
This is the airstation to fill the bottles with air from a large diving bottle.
I did continue the development of the electronics of flame tank V1.0
This one had a screen and some other electronics which on the new MCB controller will be removed.
The old MCB controller connected to its small computer, that is a simple x86 processor which runs on 300MHZ and runs ubuntu 10.04 server edition.
The PS3 will use this version too.
This is the NEW MCB controller V2.0
It supports the following features:
3 bobines to ignite the fluids.
usb slave interface to the ps3
4 sharp IR distance sensors.
i2c bus for software updates and control of the motor PCB's.
spi interface
pressure sensor (up to 10 bar)
8 servo's
external serial bluetooth module
4 external leds
electronic compass
PCB is 45 x 55 mm
I did continue analysing the PS3
The PS3 with the PSU next to it, as I want to use a 12V power supply for the PS3. I had to develop my own (stable) PSU.
This is the PS3 PSu, as you can see it is quite large.
It generated 2 voltages 12V for the board and 5V as standby power for the capacitive buttons and bluetooth receiver.
On this picture the 12V connector is good to see, I had only one problem.
I couldn't find the connector anywhere to buy, I think I will solder my PSU directly to the PS3 PCB.
And this is the inside of the originale PSU, as you can see It is a simple design.
And this is what I came up with, the specs:
buck boost design with a voltage input range from 7 to 16 volt DC
output is 12v stable and 5v constant (it can be disabled through the internal processor)
The design can handle up to 200 Watt.
The PSU can be controlled through i2c
processor updatable through i2c
voltage and current measument, these values can be read constantly.
PCB is 86 x 45 mm
PSu can also run natively (without processor in the PS3)
Next step was to look at how the track would fit.
The motor fits perfectly.
The concept of the current tank has been changed quite a bit compared to the first tank.
With the first tank a Motor controller which could controll up to 8 motors was used (through i2c)
For the new concept one processor for each motor is used, still controlled through i2c.
In total 4 wires run into the track (GND, VCC, SCL and SDA)
I want to mount the new controller directly to each motor.
(this is quite a challange as the pcb needs to be small)
This is the track,it isn't welded yet but fits perfectly.
The designed motor print has the following specs:
The design withstands voltages up to 16V
short circuit proof.
The design can controll motors up to 3 amps.
The PCB can be controller by an RC controller (as motor or servo)
digital (puls) and analog(trim resistor) feedback sensors
power measurement of the motor
PCB van controll a motor normally or as a servo, when a second print is mounted it even van controll stepper motors
PCB has a diameter of 26 mm
software is upgradable by i2c
This is the bobine used to ignite the spiritus fume.
I bought the camera at dealextrem, I am using 4 of them.
THe GPS module is also bought from deal extreme.
These are the tracks, milled from MDF wood.
Too make the surface smooth I use filler and then sand it down.
These are the plates after treatment.
After that a mold had to be made to place the bindol sand in.
Because aluminum is quite dense I use bolts to ensure I will not float while pouring.
The mold is coated by graphite so the sand releases the mold easy.
The graphite on the plates.
After the sand can be poured.
The other site is filled too, after that the wooden mold is removed and the complete thing is tightened.
After that the first test pour has been done, sadly it didn't came out quite right.
As you can see the aluminium didn't flow quite right.
This is due to the temperature and the consistency of the aluminium.
The tracks which are good fit perfectly
I created an improved burner, now the crucible gets way hotter.
You can see the crucible glow even after a few minutes.
This is the new result, WAY better.
After this I started on the Stainless steel hoods, I want to mold the metal myself with a hamer and an english wheel.
I cut the shape out of the steel.
You have to hammer dents in the metal to stretch it as much possible and then roll out the dents.
I roll out the dents with an english wheel, I use different rolls for different curves.
And then the steel needs to be rolled...
I made a movie from this process and removed the audio (because of the noice )
ps. I Anyone wants to see the schematics of the electronics I created, just ask.
I have been busy designing and prototyping the most important part of this tank:
The flame throwers!
As allways with prototyping there are successes and setbacks.
What I initially wanted to do is:
Blowing air past a hole to create a vacuum, this vacuum sucks in fuel which will then be blown into small particles (easy to ignite)
This mixture then has to travel past an constant arc which should ignite the mixture.
I first started with building an constant arc, I cracked open a few old CRT monitors to salvage the flyback driver from them.
As I was experimenting and researching I stumbled upon homebuilt plasma speakers A.K.A. singing arc.
These arcs are modulated by sound and can display normal audio.
Thus I created a prototype of such a speaker.
After some experimenting this is my first attempt, de sound is still a bit low but you clearly can hear the "C&C red alert's hells march" playing
Next part was the design of the flamethrower itself.
A large hole is drilled for the air-fuel mixture.
Also the back side has to be drilled.
This will become the hole through which the fuel will be transported to the vacuum chamber.
Next part is to drill the actual hole from the fuel hose to the vacuum chamber.
The drilled holes them are welded shut so no leakage can occur no matter the temperature.
These insulators will hold the electrodes for the sparc gap.
Yeehaa a nice constant arc
After having created the inner nozzels (which control the amount of vacuum and the fuel-air ratio)
I did a few tests.
The fluids did spray very well with my design but the ac just couldn't ignite the mixture.
I did some test with a guard flame running on butane.
I bought 2 butane tourches and created new housings for them and adapted them in such a way they could be shut off with a valve.
You can see the tourches here, they are fitted in a new brass housing.
I created this holder to accomodate the tourches, the touches will be mounted at an angle for the best ignition position.
Then I welded it toghethe with the flamethrower itself.
You can see how the electrode will now ignite the butane gas, I am thinking of relocating the electrode to the top of the flame thrower so the air won't intterupt the ozone created by the arc.
You can see the thing with the tourch attached.
And the rear.
And now the unit with the initial valve attached (this changed later)
I had to create a connector to distribute the butane to both tourches, that is why I made this.
Both pieces of brass are connected to each other by soldering them together.
And now with the air connectors attached.
Next step were the actual results , hope you like them.
In this picture you can see the complete setup, more detail is shown in the movie.
I owe you an update for far too long now.
That happens when someone has too much hobbies.
Anyway after my recent HDD crash I lost a lot of pictures but I managed to salvage some of them in my backups.
But sadly I have lost a lot of pictures documenting the welding of the tank.
Have fun watching/reading!
Further development on the source code on the still old platform.
I bent an aluminium strip around the tube to cut it in a straight line with my plasma cutter.
This part is a bit trial and error, to get the look and feel 'just right'.
It has been simply tacked to allow reloaction if is doesn't work for me.
The steel has been hammered till it was bent like this.
After that it had been rolled by and english wheel and cut down to size.
I cut out the sides by tracing it over and then cutting it out.
After that I bend it to its shape and then welded it down.
The tacks are good to see on this image, a small part has been completely welded.
I welded some reenforcements to the body to make it more rigid, also the hood and cylinders are mounted.
As you can the it doesn't fit yet like it should.
This is the back side.
The space between the cylinders is too small, I will make some provisions in the hood to compensate for this.
The front side is mounted and welded.
A picture for the overall look.
A front side view with the blue ray placed at its position it will come.
After I desided on the blue ray's position I welded its supports to the body.
Aditionally this provides a rigid body.
This is how the blue ray will be mounted.
And it even fits under the hood still
I want the cap to be removable, that is why I welded strips to the body on which the cap can be bolted.
The tack welds can be seen clearly while it is beeing shaped to the right position.
An overview of the welded strips, this makes it quite rigid.
The cap has been rounded slightly.
Next step was the mounting of the cap in the middle, this cap cannot be removed because it needs an fixed position needed by the blue ray player.
A side view of the tank.
This is how the rear hood is mounted, I will weld a piece of rod to make it fit perfectly.
An opening was needed for the blue ray player for the cd's, this has to be on exactly the right height.
The edges are kind of sharp, this needs to be solved because it will damage cd's.
Yes it fits!
I took a tube of brass and grinded a thin slot in it.
After trimming the brass I clamped it to the body, now the cd's are safe.
I soldered the brass to the body with my hot air soldering station.
The welding of the rod can be seen here.
I made some space in the hood needed for the cylinders this will be made on 2 positions to make it symmetrical.
I also made a small image of the TIG welding:
I wanted to make a screen in front of the tank where I could change the settings for the tank.
I recently changed the design but still wanted to show it to you.
It incorporated a flap which could slide in front of the screen to protect it.
I worked hard on the tank recently and thus I am again ready for another update.
I am also working on the software but it whould be rather boring to show foto's of this proces.
With the thicknesser I shaved an MDF plate down to the right thickness for the tube sealings.
After cutting the plate into small rouds I sanded them down.
The top aluminium cover needs grooves to get neatly placed on top of the tupes.
The mounted bearing on the mill prevents you from cutting too deep into the wood.
I clamped the mill into an vice, now I can easely cut the grooves into the wood by rotating the round plates around the mill.
It fits perfectly...
On this picture it it more clear on how it will look when the plates will be casted.
I will cast the plates with aluminium, I created the sand molds with the MDF plugs, next up will be the casting itself.
Next part was the creating of a bearing mount, the top part of the tank will rotate on the bearing, much like a real tank.
The bearing will be mounted in this way and secured by tightening a bolt.
Sadly the MDF distrupted my sand casting, this had to be repaired with my mill.
And here the end result.
I drew the mounting lokation on the bottom plate.
Then I bolted the bearing mount to the tank and cut a hole with my plasma cutter.
Through this hole the cables will be installed...
12v, HDMI, audio etc..
Also a motor will be attached to turn the top part.
Finally my PCB's arrived last week after "only" 223 days of waiting
I immediately ordered all the components:
I have soldered and tested 2 PCB's up till now:
The bluetooth pcb and the main controller board pcb.
First I warm the pcb up to a temperature of around 130 degrees celcius, after that I bring the hot air soldering iron close to the chip for about a minute (till the chip solders itself to the pcb.
Naturally I use some no-clean flux, this makes the solder flow better.
an indication of MLF technology (this is an 20 eurocent coin)
First I attach solder to the chip and after that I relow it to the pcb, I let gravity to the hard work for me.
Also here I am warming the PCB before reflowing.
This is the main controller board, tested and ready.
The PS3 power supply is done and all works in the first try.
Up till now I only found one small design error (this was easely corrected bij removing a small wire).
Below You can see the bottom, the diodes and mosfets stick out significant.
And here the top part, This PCB is completely self sufficient.
But I still added a small processor with I2c bus, this controller monitors the power usage of the PS3 and can shutdown and start it up when needed.
Also the main controller board can do these actions through this controller.
I have tested the stability of the design from 8-24V with 0-2.5 Amps, no problems have been found.
The signal is and stays perfectly stable with hardly any ripple.
(I can't go higher then 2.5A with my LAB power supply)
Picture of the ripple at 10V, 0Ampere measurement range: 20mV per decade, frequency of the pcb is now 389KHZ and the average ripple is 0.054V Peak to Peak.
Picture of the ripple at 10V, 2Ampere measurement range: 50mV per decade, frequency of the pcb is now 397KHZ and the average ripple is 0.148V Peak to Peak.
C&C tiberium wars Advanced flame tank
)
)
