Sunday 28 August 2011

Buren and I worked yesterday on fixing up the oxidizer tank mounting. I was thinking the other day about the run tanks and concluded that their no different from the tanks used by commercial hybrid rockets (as the tank is of course vented to reduce the pressure inside).We decided to just bolt the tank to the test stand as we were having difficulty with the steel round system.

We also removed the actuated valves and mounted the valve hardware to a much thicker piece of c section as the valves had a bit of flex at the last test. Also at the last test Buren broke off the head of one of the bolts holding the valve plates on so we had to drill out the remaining pieces of bolt. They came out easily so the hole didn't need re-tapping. I also made the mounting plate to hold the potentiometers (for feedback).

I am just waiting for the electronics to come then I can start on upgrading the electronics.

Unfortunately the next test will have to be delayed by a week because Ariel and Rob will be busy during the planed test day.

Thursday 25 August 2011

So the long network was more problematic than I had initially though (in that it doesn't work). I took a spool of 200M of cat5e cable, crimped two ends and connected my computer to network and at first it looked good (flashing lights) but it would only stay connected for a a few seconds then would drop out. I tried forcing 10mbit and it would stay connected but i was unable to log into the router or ping any other computer. I thought of trying with cat6 but didn't really want to buy a box if i wasn't sure it was going to work. I guess I will have to use wifi if i want networking to the stand. I am going to try some range tests with wifi and see if I can get 300m range from a router wit default settings/antennas. If I can't i will have to go to directional antennas. I still think a wifi network will be more reliable than the xbee just because a nework has allot more redundancy built into it. 

The USB network server arrived the other day and works really well. I thought it might not work too well with a usb-serial chip but it worked first time which I was really happy about. The software is quite basic (which is good) and after installing it I was able to connect the FTDI chip on the Arduino to my computer and program over the network. Another feature I like is that different computers can connect to the usb device (only one per port) so we can have a few people monitoring the different webcams while I concentrate on the control program. 

I decided to go with proper motor controllers for the valve motors the other day (over relays) as it will allow me to have much more control over them. Previously i was only able to send about a 5ms pulse which was about 1/8th of a turn but with the controls i won't even need to pulse the valves as I will be able to vary their speed. Eventually i hope to have a control system using the feedback data to actuate the valves so all i will need to do is select a position and the valve will go to it. I ordered a nice 2A dual motor controller. The other advantage is everything will be much neater as I eliminate 4 relays and the controller is quite compact.

I have also been working on the safety operational plan. Writing it was quite a daunting task as it needs to cover allot but I started with some checklists and have been building from there.

We have also secured a slightly closer test site (1/2 hour drive). If all goes to plan I would like to have everything finished  by next weekend for a test the following saturday or sunday. Last time we were rushing to get everything finished and this time I would like to have a week to clear my head and double check everything is %100 operational.


Monday 22 August 2011

The last week I have been busy catching up with uni work so I haven't had time to do any rocket stuff, but I have ordered all the parts I need with long lead times.

I decided to replace all the relays in the relay box with a nice neat relay bank as the transistors on the old ones are dead (not too difficult to replace, but i still haven't identified the cause) and don't want to rick the same problem. The new ones are optically isolated from the micro controller which is a nice feature. Also the igniter will be on a separate battery.

I have also decided (reluctantly) to ditch the wireless control scheme and go for wired. I really liked the wireless control system and haven't had any major issues with it, but it is not %100 reliable (does loose connection from time to time, although doesn't affect operation) and after the last test I have come to realize everything has to work %100 of the time (not %99) because if something can go wrong it will. Wireless does have its uses, but I think a cable is better for this application.

The new control scheme will use a 300m length of cat6 cable, through which I understand I should be able to maintain at least 100M/bit speeds. There will be a usb-network server on the test stand which the micro controller is plugged into and shared over the network to a computer. The other advantage of a cable is that I will have enough bandwidth to run streaming video from the stand. The usb server supports 4 devices so I plan on having one usb tilt-pan camera and another fixed webcam. This should give me a good view of whats going on with the stand. One concern I have is that the usb server won't work well with the Arduino, bit I have ordered one and will know soon. I can always use a computer on the test stand sharing usb, but would prefer not to. I had considered a more direct approach of using a arduino with a lan port, but I really didn't want to go into programing for that and although the hub adds one layer of abstraction (and source of failure) it will act as a serial link (same as the xbee) so I won't need to modify the control program at all. Also if the cable is struggling (which it shouldn't) I can uses booster.

Valve feedback is the other major thing that need implementing. After some discussion with ariel about the different options (trimpot, 12-way encoder and optical encoder) I deeded to go with a simple trimpot. I did find a nice 100 count encoder which was relatively cheap and would be more accurate but the problem is its not absolute and could need calibrating. I thought of implementing an auto calibration with two limit switches but why complicate something if you don't have to, and with a %5 trimpot i will have all the accuracy I can handle.

Also I need to add a manual vent valve to the nitrous side as after I lost control of the stand we had no way of venting other than manually turning the throttle valve. I was thinking of attaching a rope to the two manual vent valves so that in the event of a total loss of test stand control I can still manually vent from a safe distance. Again I could go for a complex redundant electronic backup system (that vents electronically) but why complicate things! Also I want to remake the valve hollers out of thicker aluminum C section, as I noticed some flex in the 3mm one.

Much work also needs to be done on the operations manual. I really want it to be something that is useful, not something that is done because it needs to be so I plan on compiling it over a few weeks so I can give it good thought.

I plan on working with Buren on sunday on the mechanical things, and we should be able to get most of them done. We are hopping to be ready for anther test in the next 2-3 weeks or so. I also need to buy a markee, table and chairs before the next test. I was tempted the other day to get a new generator (11KVA petrol for $1000!) of ebay, something I have wanted for a while but don't really have the need for at the moment Using all my power (pun intended) I managed to avoid becoming one of those people who buy tools not because they need them but because they want them (which is something that annoys me) With a surge protector (so it doesn't ruin my charger again) my 500W  one will be enough to power all our devices. Also  at the last test Marco suggested I have something fun I do before each test (the Russians used to pee on their rockets for luck). I have thought of a good pre-test ritual, although I am not sure everyone else will enjoy it as much as I will - hint - we have a somewhat different taste in music......

Tuesday 16 August 2011

The first of manny tests

We conducted the first test on the weekend and it was successful, and although I didn't get all the tests I wanted to do done due to a technical issue with the relay box overall it was successful. As I will discuss there is much to improve in the way of safety and organization.

I wanted to do three tests on sunday:

1: Filling the nitrous tank remotely, and venting it using the actuated valves.
2: Filling the fuel tank, filling the nitrous tank remotely and then venting and igniting the propellants
3: Same as #2 but with combustion chamber on (very short duration test)

Test #1 went as expected. After opening the nitrous fill bottle valve and retreating I opened the vent valve, fill valve and then closed the vent after about two seconds, followed by the fill. I then opened the oxidizer valve and Hissing was heard as the nitrous vented through the injector block. One thing that became very evident was that it is very difficult to keep track of the valve position in your head, especially in the heat a first test. I had two buttons, one to open them 1/10th and one for full but between watching and listening to try to see what was happening I completely lost track of how manny times I opened the valve. Not a really big issue as it it can't go past open, and to be sure they were closed I closed them a few times but for the next test I really need to have feed back on them. Just a potentiometer, coupled to the motor shaft will work well.

Test #2 started well, I filled the fuel tank roughly 1/4 full of ethanol, opened the pressure (argon) and nitrous bottles and retreated, after arming the igniter which was installed just past the injector block. As with test #1 I filled the nitrous tank. I then pressurized the system then ignited the igniter. After seeing that the igniter did ignite I opened both fuel and oxidizer valves but nothing seemed to happen. I was still getting confirmation that the mirocontroler was receiving commands but the valves (and nothing else) wouldn't respond. After waiting a bit (not long enough) we approached the stand and found hissing from the top of the fuel tank (the vent line at the flare connector to tee adapter) and in the female-female adapter on the top of the ox tank. We did pressure test with 120PSI air after setting up and didn't find any leaks but the lines probably came slightly loose during the bumpy drive. After the tanks had emptied (I manually vented the fuel tank, and we waited for the ox tank to vent) I could all the sparkfun type relay boards to be unresponsive. The ones rob made with the DPST relay and reed really were working, so I could close the valves but not open them and nothing else worked as they all used the spark fun boards. I haven't had time to diagnose the problem, but it seems like the transistor that switches the relay had blown on every board. I don't understand how this could happen, as they are all wires in parallel, but ariel thought that the ignitor could have acted as a big inductor and fried them. This is consistent with the fact that everything stopped responding just after the ignitor lit. Unfortunately I don't have the burnt ignitor to see if it burnt through, or if the relay stopped working before it got enough power to burn through. Marco and I worked for about half an hour trying to diagnose the problem while burn and ariel fixed the leaky pipes but we couldn't fix it and it had started to rain so we decided to move the trailer back to the main road and try to fix the box in our car while the rain passed. I couldn't so we packed up.

Although I had a rough idea of the things I wanted to do on the day of the test I didn't have a set plan, and I   learnt a really important lesson about the complexity of running a test. I future I would like to have a plan of the test in detail including different scenarios like the weather, failures etc. The weather coindincentally was something I had given no thought to and when we saw rain coming we just said....well better hurry up, which was entirely the wrong decision because we were then rushing. Organization of tools and parts was another area for significant improvement, as we misplaced tools manny manny times. I hadn't given any consideration to balancing the trailer and we ended up using a spare argon tank to weight the trailer, which one time in the rush before a test I removed, only to have the trailer tiped over (i cut my finger quite badly In the process which was very distracting when it came to the test). Rushing I think was the thing we did the worst, and I think this was a combination about not really having a set, rehersed plan, the weather and being generally nervous and wanting to get it over with.

The main safety issue I identified was when the test stand stopped responding, and we approached the stand with the tanks partially pressurized (although not fully because of the leak). For this I need a backup remote venting system. This can be as simple as string attached to valves. Test stand stability was the other. The other thing was fateuige (of us). I had thought of every possible spare tool, spare part and piece of equipment to bring (preparation was one thing well done) but the things we needed: Food, water, sunscreen I neglected. I was getting really dehydrated towards the end of the day because I didn't bring any water, and I think everyone was getting a bit grumpy because of the lack of food.

Also the drive was long and quite rough, although the test stand held up quite well accept for the two leaks. In future I would like to do a high pressure test on site. I was hopeing to avoid this because it uses a lot of argon with the tanks empty, but is necessary. We did have an issue with one of the tank mounting points cracking during the drive (the scrap piece of metal I used was apparently cast iron), but we just strapped it up and everything was fine. Both these will need to be replaced before the next test.



Technicial issues which need to be resolved:

1.Key in fuel valve needs to be thicker as it fell out and we had to apply a temporary fix of tape and plastic.
Valves need feedback

2.Need to install pressure transducers (already have them) so tank pressure can be confirmed as empty before approaching stand

3. Valves need feedback.

4. Need to have a isolated igniter circuit, with separate battery so it can't interfere with the control circuit. Would also like to have a backup ignitor system.

5. backup venting system

Organizational issues:

1. Need to have a written operations plan for the test that includes:
-A procedure for the setting up - With checklist
-A procedure for pre test, test and post test
-A procedure for dealing with weather, and unexpected outcomes (all of them!)

2. Need to rehearse the test before hand so everyone knows exactly what they need to do

3. Need to consider other test aspects other than just technical.

4. Have a extra table for tools etc at the test site and things neatly organized (rushing didn't help)

5. Not rushing

6. Get a good nights sleep!

Hopefully we can get a test site that is closer for the next test. Driving slowly with the trailer it took about 2.5 hours just to get to the site. The drive alone was quite taxing. Andrew suggested I try talking to some people with more experience to develop the operation plan, which is a good idea.

The videos of the actual test are quite uneventful, and all you see is some lovely clouds in the background accompanied to loud hissing. I won't bore everyone (unless they want to see it) with the video of the two tests but here is a quick video of everyone setting up.





I have been a bit critical, but overall I would say that the test was a success as I showed that I could remotely fill the nitrous tank and vent it using the actuated valves. Considering that this is the first test we have done on this scale and complexity I think everyone did really well, and I would like to thank Buren for all his help over the last 8 months. Also tanks to Ariel for his help through countless conversations about electrical aspects ( I would ave been using either a really long lan cable or a wifi antenna with directional antennas is it went for you!) and for his help on the day. Stewie you were also a grab help with the relays and strain gauge amplifiers it would have taken me quite a lot longer to understand the amplifier circuit. Also thanks to Marco, (who I can say was probably the only person thinking straight on the day) for his help on the day and help and advice and encouragement over the last 6 months. Thanks to Andrew for your help, advice and encouragement.







Sunday 7 August 2011

I was debating weather to bother with making the new tanks as its allot of effort, but i am very glad I did. Also I should mention that my friend Mitch came up with the threaded rod idea.... thanks Mitch! Buren and I spend yesterday and today making and mounting the new tank the the stand. When we opened the old one we found allot of congealed grease mixed with rust all over the orings and the end of the tank. It would have not ended well if we filled it with nitrous. The new tank actually didn't take long to make but mounting it to the test stand was quite a hassle as we couldn't get it secured well enough and it would always come loose. I am quite happy with how the tank came out, and it is completely severicable so in future we can just open it up to clean it instead of having to cut the pipe up to get the caps out.











We also started cleaning out the nitrous side lines with acetone but the tanks took one day longer than expected so we didn't get that finished. All that needs to be done before the test next sunday is finnish cleaning out the lines and sort out the valve timing (time required to throttle and fully open the valves).

Tuesday 2 August 2011

Updated program

Did a bit of work on the control program tonight. A few of the buttons are not functional yet, but everything is how I would like it for the first test. Possibly will add in a display to show the approximate valve positions.

Monday 1 August 2011

First test date set, odds and ends + new tank

I have been busy with the start of uni for the last week, but also have been talking to a few people about test sites and also some of the safety aspects to do with working with nitrous oxide.

We have found a location to conduct the first test, and set the date for the test as the 14th. Thanks to all the people who suggested test sites!

Things I need to do before the first test:

Mechanical:
Make 10 igniters
Buy acetone and clean nitrous side lines
Dry lines
Modify plumbing for new pressurisation scheme
Make line from argon regulator to argon inlet
Flow test @ 500PSI - Check for leaks
make new tanks and install
Wire up argon solenoid
complete dry run
Clean relief valve with acetone
Set relief valves to 900PSI?

Following a discussions about nitrous oxide safety issues I am concerned about the traces of oil that might be in the main tanks. I have washed them out throughly, but There could still be oil trapped behind the o-rings and any oil would be really bad. The problem is that the current design does not allow for opening (have talked about this in detail before). I pressure test with oil, so i cant pressure test, then clean it internally. I have a new design which involves not bolting to the pipe, but using lengths of threaded rod running the length of the pipe to take the axial load. I have all parts required and if anything it should me much quicker than a regular tank because there are only 8 holes to drill (vs 12) and I don't have to thread the holes. One thing I am aware of is fracture induced by the thread in the rod, as it will introduce a large stress concentration. I have calculated this, and will also pressure test.

Software:
Add in pulse right forward + reverse buttons
Add in on for time buttons

The software is still rather basic, But it will do.



After I run the first test I will have a better Idea of what features i want in the control program. I have ideas for a automated sequence and other advanced features, but i want to do a basic test first to make sure everything works before I complicate it. One feature I will need to add before the test is button that moves both valves to low throttle and another for fully open. Also a close valve button wouldn't be a bad idea.

Also for the first test I won't have time to install the pressure transducers or have a working load cell amplifier but I am more concerned about other things at the moment.

A few things to do but I have till the 14th.