Saturday, 15 October 2016

Oven Ready

Last week I finally got around to testing the pottery kiln that I intend to use as a heat treatment oven.

The tube bundle thrust chamber will be a weldment composed of 6082 and 6063 elements. In order to solution treat or artificially age this I have to raise it to a temperature of 525 degrees Celsius +/- 10 degrees and hold that for a specified time.

The kiln is fitted with what looks like a relatively sophisticated temperature control system based around a thermocouple. I doubted whether this would be very accurate given it's intended use for firing pots. The temperature control dial is calibrated in 20 degree increments, so I wanted to find out just where exactly 525 degrees would lie.

The top and side of the kiln are fitted with inspection holes that are plugged up with ceramic bungs. So I draped one of my K type thermocouples into the top hole and carefully replaced the bung so as to trap the cable in place. In this way the thermocouple was suspended roughly in the centre of the kiln's heated space.

I had the thermocouple connected up to one of the amplifier units that I have shown previously. The output from this was fed to one of the ADC inputs of an Arduino Uno. The Uno was programmed using LINX toolkit for LabVIEW. The LabVIEW application that was built for the Uno was running on the laptop to which the Uno was connected by USB link.

The workshop roller door was fully opened to guard against any build up of fumes and a fire extinguisher was placed close at hand. Having set the temperature dial to what I reasonably hoped would give 525 degrees, I plugged the beast in and fired it up.

A satisfyingly loud hum emanated from the kiln and the temperature began to climb almost immediately. At around 200 degrees smoke started to escape from the lid of the unit, which at this point was slightly ajar. I grabbed the extinguisher in trepidation but I needn't have worried; clearly just grease or some other residue burning off after months of disuse.

Satisfied that all was well I shut the lid properly and snapped closed the toggle fastener to make a tight seal between the lid and the main case of the kiln. The temperature was still increasing steadily and the rate of increase had gone up with the proper closing and sealing of the lid.

After 25 minutes of running the 450 degree mark was breached and after 35 minutes a temperature of 520 degrees was attained. This finally settled down at 530 degrees and stayed there, give or take a degree either way. 

I didn't think this was too bad at all, what with my initial misgivings about the accuracy of the control system and that the dial had been set quite roughly.

Below is a screen shot from the LabVIEW temperature display showing that 531 degrees Celsius had been attained.

More details to come on the tube bundle design and tube bending arrangements.

Sunday, 2 October 2016

Thank you Iain!

You all know that the subject of this blog is amateur rocket engineering. I've always thought of myself as the most amateur of amateurs, but I'm pleased to report that my small attempts have garnered attention on the other side of the world.

Iain Finer is in New Zealand and is therefore my Commonwealth cousin, as well as a brother in the fraternity of rocket engine constructors. No slouch in the engineering department, Iain has completed an excellent regeneratively cooled liquid fuelled engine, shown below. 

Not content with his own masterful efforts, he spends the rest of his time reporting on all things reaction related, i.e. what the rest of us are up to.This he does through the medium of his superb blog "Mach 5 Low-down":- 

Iain has very kindly picked up on my tube bundle concept and has added British Reaction Research to his list of featured blogs. I have to confess that I am getting a nice warm glow from the knowledge that I am now up there with all those luminaries. Then again, it could just be that I have left the heat treatment kiln on again...

Here is Iain's completed engine, as promised. He is currently in the middle of building a test stand for this beauty, and I can't wait to see some mach diamonds:-

Thanks again Iain for your interest in my project. Now I'll have to produce something worthy of your august pages...Stay tuned.

Sunday, 4 September 2016

Patently Obvious

My internet research frequently involves trawling through old patent documents. These often provide a wealth of information, and I'd like to share some of this with you now. 

Great Britain patent 376,974 was applied for in August 1931 by the Swiss engineering firm Brown Boveri. Granted in August 1932, it has as it's object "Improvements in and Relating to Combustion Chambers".

This is one of the earliest references I can find depicting a combustion chamber made from tubular elements to allow a coolant to be circulated through.

Whilst the invention described was intended for the production of power, not thrust, it is easy to see where the concept of a tube bundle thrust chamber may have originated.

It is highly instructive to quote directly from the patent document itself:-

Reading further we find:-

This is exactly what I am trying to achieve. Some of the drawings accompanying the application are strikingly similar to my own sketches, and the constructions shown are mirrored closely by my earlier tube welding trials:-

The document also describes the use of bands of material welded around the tube bundle to take the hoop stress caused by the combustion pressure. All of this information will go in to the pot and add to the tube bundle design effort.

Wednesday, 3 August 2016

Dyne-ing Out

Greetings all. What's this title all about then? Well, the dyne is a unit of force, and I am away from home...I am back out earning a living at present, so this is a bit of a first for my project and the blog - the company I work for have allowed internet access to personal sites and so I'm able to post here whilst I'm away.

I have been looking into the forces required to break a hypothetical aluminium test piece, in order to come up with a set of dimensions that keeps the pressure needed within acceptably safe limits.

The hose on the enerpac type ram I'm using is rated to 10,000 psi. That said, the fittings that I have to connect a gauge in the system are only good for 6000 psi. So there is my limit.

The 6082 plate I intend to make the test pieces from is 6mm thick. A 10mm wide gauge portion of this material would thus have a cross sectional area of 60mm square.

As the average UTS of 6082 T6 is 300MPa, or 300N/mm^2, the force required to break a piece of this area is 18kN.

18kN translates to 1835 kgf. The surface area of the hydraulic piston is about 7 x 10^-4 metres square. As pressure is force divided by area, this gives 2621428 kgf/m^2.

In more familiar units this comes out to about 3730 psi, or 257 bar. This is well within the acceptable safe range for the fittings and hoses being used.

This time out my work has involved a project with an aluminium TIG welding element. Here is a photograph of a joint I made yesterday. The test pieces will be cut from a plate that has been halved and then welded back together again with a seam similar to this one. The control piece will of course be a solid section, that we now know will take about 3700 psi to break.

What figure can we expect a newly welded, non heat treated test specimen to break at? Assigning a value of about 190MPa as our UTS, using the calculation method above gives a force of 11.4kN. This translates to 1162.5 kgf. To produce this using the enerpac type jack will require a pressure in the region of 2360 psi or 163 bar.

Theory is one thing; practice can be downright embarrasing. We'll just have to wait and see.

Sunday, 26 June 2016

Aluminium Tubes

As a material for thrust chamber construction, aluminium is superior to stainless steel in a number of important regards.

Steel is typically 2.5 times the density of aluminium, which has obvious impact on power to weight ratio. Aluminium has 92% higher thermal conductivity compared to stainless steel, easing cooling circuit design. Although aluminium exhibits lower strength at elevated temperatures, this tends to be offset by the material’s excellent heat transfer characteristics. Aluminium is easier to machine and less expensive, enabling time and cost savings to be realised.

The strongest of the commercially available 6000 series alloys is 6082 T6, with a UTS in the region of 300MPa. It can be had in all the major sections, as well as plate and sheet. That said, the smallest tube diameters tend to be on the large side for the projected design application. For suitable tube it is necessary to use 6063 T6, which is ideal for these sizes as it is easier to extrude. This has a UTS around the 250MPa mark.
The graph below shows the behaviour of various alloys, including 6082, at elevated temperatures (European Aluminium Association).

As shown, 6082 maintains 90% of it’s relative strength at 150 degrees Celsius. The other 6000 series alloys follow this relationship closely. The graph below shows the effect of exposure to high temperatures on the room temperature tensile strength of 6061 T6 (MIL HDBK 5H).

At 220 degrees Fahrenheit (approximately 100 degrees Celsius) the tensile strength of 6061 T6 shows no degradation, even after 10,000 hours exposure. This should be familiar territory to anyone with an aluminium saucepan.

The sweet spot temperature wise for the 6000 series alloys can be seen to be 100 – 120 degrees Celsius. Providing the metal starts out in the T6 condition it should stay that way and have sufficient strength to withstand the combustion forces.

The high thermal conductivity and high emissivity of aluminium makes Twg figures in this region feasible, given the current fuel/coolant design flow rate. Basing the chamber wall thickness on 90% of the UTS of 6063 T6 gives a factor of safety of 1.3 at twice the design chamber pressure.

The main drawback to the use of aluminium in the present scheme is the loss of temper brought about by welding. The properties of the section in the weld are lowered to the T0 condition. This is a loss of tensile strength of some 80%.

However, aluminium weldments may be brought back up to the T6 condition by post weld heat treatment. This typically takes the form of solution treatment followed by artificial ageing.

It can be seen that the advantages accrued from the use of aluminium as a material for rocket motor construction make research into the heat treatment problem worthwhile.

The table below is taken from a Welding Research Supplement by the UK Welding Institute from 1983. The subject of the research was the effect of post weld heat treatment on 6082 T6 and 5083 plates and sheets. The plates were cut and then welded back together, with test pieces being taken from across the weld plane. A typical test piece is shown below the results table. The 6082 T6 sheet was TIG welded with 4043 and 5556A rod to investigate the effects of the different fillers. 

And the test piece dimensions:-

The table shows that in the as welded condition (AW), the 6082 plate joined with 4043 filler showed markedly lower 0.2% proof stress and tensile strength. After artificial ageing (AA) the figures are improved, but still significantly lower than the T6 condition. Both these welds failed in the heat affected zone (HAZ) and show limited elongation. 

By contrast, the 0.2% proof stress and tensile strength of the solution treated and artificially aged (ST & AA) sample shows values much closer to the un-welded T6 condition. The elongation has also improved, and the test piece failed in the base metal (BM).

The table portion below comes from the UK Aluminium Federation's publication "Strength of Aluminium Alloys". This shows the temperatures in degrees celsius, hold times in hours and quenching medium required for solution treatment and artificial ageing of both 6082 and 6063 alloys.

 So for a tube bundle motor weldment composed of 6063 and 6082 materials, 7 hours at 525 - 530 degrees celsius followed by a water quench and then 12 hours at 170 - 180 degrees celsius would be called for. 

Commercially available heat treatment ovens are prohibitively expensive. That said, a pottery kiln is capable of achieving the required temperatures. I have managed to get hold of a second hand kiln from ebay. The unit is able to reach a maximum temperature of 1200 degrees celsius. A thermocouple within the heated space enables closed loop control of the temperature. The internal volume is 300mm^3. Here is a photograph of the kiln:-

I intend to produce a number of test specimens similar to the one shown above. These will be in 6082 T6 plate. The test specimens will be of three types, welded with no heat treatment, welded and heat treated in the kiln, and finally a solid section of plate with no weld as a control.

In order to subject these pieces to a tensile force I have designed and am in the process of constructing a rudimentary tensile testing rig. The next three photographs show the rig so far:-

The rig is composed of 50mm x 50mm x 4mm steel box section. The test piece fits through the two opposing slots and is secured by pins through the cross drilled holes. The M12 high tensile threaded rods pass through the rectangular frame and terminate at a piece of box section. An enerpac ram will bear against this box section. The ram provides the force to pull the test pieces apart. 

Force will be inferred from a pressure gauge and extension from a DTI. I could acquire both values by electronic means but I do not want the unit to become a project in it's own right. Based on the ram piston diameter, a pressure of 400 bar (6000 psi) should produce a force of 29kN. This will be enough to break a test piece with a gauge area of 90mm^2.

I'm hoping that these tests will show that solution treatment and artificial ageing using the kiln will work. This will be an important step forward in the production of an aluminium tube bundle chamber.

Thursday, 14 January 2016

Alright, Chuck...

Greetings all. If you are not from the UK, and specifically from the North of England, you are going to need an explanation of this post title. "Chuck" is simply a term of endearment, which is actually a contraction of "chicken"...but enough lexicography.

I have almost finished mounting the ER32 collet chuck to the backplate (see what I did there?). All that remains now is to drill the mounting bolt holes to join the chuck and the backplate together.

After mounting the D1-3 camlock backplate, I machined a register on it that is a very good fit in the recess on the back of the collet chuck. The collet chuck was then fitted to this register, ensuring a very high degree of concentricity.

Tony at has written an excellent primer on the fitting of chucks that can be found here:- I found this to be very helpful.

Here is a photograph of the collet chuck mounted on the backplate:-

 You'll see that the chuck has three bolt holes. These are tapped for M6. I cannot put bolts through the backplate into these due to the camlock studs. So I am going to drill these out to M6 clearance and then tap the front face of the backplate. The attachment bolts will then thread into this from the front of the chuck.

Machining cast iron is a very dirty business, as you can see from this picture. The swarf takes the form of fine chippings. Needless to say these can be potentially very injurious to the lathe. I managed to minimise the amount of debris by attaching the workshop vacuum cleaner hose to the toolpost. Had I not done this, my poor machine would have ended up buried.

The purpose of the collet chuck is to hold the tubes whilst the ends are being machined, as per previous posts. With the chuck mounted on a backplate, the full length of the tube can protrude through into the spindle bore. I may need to make a push fit insert for the opposite end of the spindle to act as a guide. This will stop any tendency of the tube to whip whilst the spindle is rotating.

This is all part and parcel of making ready to trial the tubular chamber fabrication method. Once I have the chuck finished I will need to look at making a system of stops for the lathe. This will hopefully ensure that all 60 tubes end up the same, within reason. I will also need to look at a means of uniformly bending the tubes, Again, this will involve a series of stops used with the tube bender.  

All of this has to be fitted in and around normal family and work life. So progress tends to be slow, though steady. Thank you for your patience and your continued interest in my project.

Sunday, 10 January 2016

Aspiration and Inspiration

Apologies, gentle reader, for the recent dearth of posts. I have been rather busy of late with non rocket engine related activities. However, I have managed to get a few hours in the workshop here and there. 

A brief summary of practical activities that have been progressed:-

Cutting jig completed and adjusted to cut square

New ER32 collet chuck in process of being fitted to backplate

I am also in the middle of writing an article on the automated indexer for the British magazine "Model Engineer's Workshop". If you are in the UK, do look out for it.

As for the aspirational and inspirational content of this post, read on...

I expect you are probably wondering what I am going to do with all of those tubes. Well, here is the photograph that provided my original inspiration:-

This frankly beautiful little tube bundle engine was built in 1970 by Maschinenfabrik Augsburg Nurburg (MAN) of Germany. I discovered it on the website of the Deutsches Museum. Sadly, there is very little information outside of who built it and when. My best guess is that it was some sort of technology demonstrator, and is of brazed construction. There is no information regarding what the propellants were or if it was ever fired.

The next two pictures come from the Smithsonian Air and Space collection in the US:-

This experimental thrust chamber was made by Reaction Motors in the US, in about 1947. It is reckoned to be the first attempt at tube bundle construction. I like this photograph because it shows exactly what I want to do. You'll notice that the tubes have been crimped or "booked" at the throat. According to the information provided, the construction is welded stainless tube. Obviously in a device that was intended to be fired the welds would need to run the full length of the tubes to provide structural integrity. Strengthening bands would also be required, like this:-

This chamber was built by Aerojet in 1948, and apparently was arrived at in isolation from the Reaction Motor's development. It is a much more complete effort, with manifolding at the head end and strengthening banding. Just look at the tube bending that has been employed to turn the fuel back towards the head end at the exit; it is almost like basket work! Also note the booking of the tubes at this point. There are no details of construction but I suspect this is brazed. Oddly the brazing does not appear to extend to the full length of the divergent section.

I would really like to learn more about these motors. If anyone out there has any more information I'd be glad to hear from them.

Progress is slow, as always. Finding these images has spurred me on though, and when you can see your goal it is generally easier to achieve it. Keep checking in for updates, and thank you for your continued interest.