What I have been doing (Novice endeavours with rockets)

[Baldor]

Jman, horizontal and you can forget the weight of the motor. I should make some tries after the first run of tests is finished.

 

Making a stand is easy. Just make a heavy and stable base for the load cell, connect an openscale to the cell, and the openscale to the PC. Use your favourite terminal software (I use termite in windows and moserial in Linux) for capturing the data. Then you import the captured data in a spreadshet and you are almost done.

 

I made mine with 40x30mm rectangular steel profile, 2mm thick that I had laying around, self treading screws, and some screws and nuts I had. I only bougth the load cell, the openscale and the USB cable (openscale still uses miniusb). Even the capturing laptop is an small laptop donated by my nieces, who no longer use it. If you have some basic tools (A saw and a drill), you can make the stand in one morning. Look at my first post for an image of the stand, and ask questions if needed.

[stix]

The reason I tare before is to see the weight of the motor before and after the burn in the collected data. This way I can compare with the weight measured in a precision scale. Once the data is in the spreadsheet it's easy to "tare" it again. It should also make easier to interpolate the fuel weight once I decide to do it.

 

Ok, so some questions. Does the OpenScale device register negative values? If it does, then you could tare to zero at the start and use the negative value at the end as the fuel consumed - it should be correct. I had a look at the OpenScale device and it looks like a simple and effective device, I really like it. My only bitch about it, is that it is limited to the highest sample rate of 80Hz (80 samples per second). Also what resolution is it capable of? It looks like from your data supplied, 1 gram resolution?, so that part looks very good.

 

My device, and issues regarding sample rates are thrashed out in this post (started about 2yrs ago):

https://www.amateurpyro.com/forums/topic/11545-thrust-metertest-rig-prototype-ver10/

 

In any case Baldor, you have the recorded data. It's how you evaluate that data, and in my view, "Fuel Loss Compensation" has to be taken into consideration for vertical devices. My basic assumption is that "measured thrust" at the time point is "exactly proportional" to the amount of fuel burnt. This somehow must be accounted for, and the final data adjusted accordingly. Small tests, especially using 30grams of fuel are probably more important to ensure that the fuel loss is compensated for - larger ones, perhaps not.

 

 

Jman, horizontal and you can forget the weight of the motor. . .

 

I agree 100% with that. However, for practical purposes, it's much easier and more portable to make a vertical device. So again, then we simply must account for the fuel lost.

 

That's my view, and I'm sticking to it, and have allowed for that in my recording software. In the end 99.5% accurate results? - might be pushing it a bit though but easier than having to make a horizontal device. It's also very difficult to calibrate a horizontal device.

 

Anyway, it's all lots of fun and good to discuss other ideas and interesting to see other peoples "contraptions" that they make.

 

Cheers.

 

[EDIT] Ah yes Baldor, I just had a look at that old thread posted above and have realized that you were a participant at the end.

 

I think you did say that you didn't like my software or approach because it seemed too complicated. Well, it's much easier than fiddling around with excel spreadsheets etc., also my software enables you to record direct from your device. Therefore you can see the results instantly. It looks like the OpenScale device is well suited to this software.

 

My view is that when you are testing, if you can get to the results quickly, then you are in a good position to test again, quicker. Otherwise it becomes drudgery mucking around with various software, and therefore less likely to do further tests. That IS the point of my software. It's all about, and for those who enjoy testing. Simple. Although, I have to admit that some of us enjoy doing it "our way" regardless.

 

Nevertheless, If you are interested, or anyone else, I'll submit the software. The current version doesn't include altitude calculations - I'm still working on that part.

Edited March 12, 2018 by stix

[Baldor]

Stix, openscale have 24 bits of resolution, more than enough. You can calibrate with the units you like, but the software will report Kg or pounds. I just put a 1Kg weight and told the openscale it was 1000 someting (grams) Something I didn't find how to configure is removing the units from the serial output, so I have something like 1234,Kg, Notihing dificult, since it's open source, but its easier to remove the units in Notepad++

 

You are right, 80SPS is low but enough for the moment. If I find I need more sample rate, I will bit the bullet and use a PIC or an cortex. I think there are some "new" Pic16 with very nice ADCs, or a dsPIC, but this will be overkill. The problem is not the uC, but to make a very low noise amplifier for the cell.

 

30g is about 0.3N. With more than 60N peak and 18.3Ns total impulse, is negligible. If I was trying to squeeze the last Newton out of a design, yes 30g is important.

 

I like a lot your software, don't get me wrong. What I will not like is to have to develop it. :-) I'm no strange to microcontrollers, I did some developing for my other hobby, photography. IR break beams for birds, time lapse, drops of water.... Just I'n not in the mod to develop from zero now. So yes, I take your offer and will test your software. :-) My workflow will not vary much, since I plan a few tests, make the motors to be tested, mount the test rig and the camera, make all the planed tests in a row, and then analize the data inside, with a big screen and some coffee. Once you have the first spreadshet, the rest are just one "save as.." and a litle copy paste away.

 

Do you have a link to the load cell amp in your stand, or some reference in the board? I will like to take a look at the capabilities.

[stix]

Will do. I'll post my software tomorrow or the next day. It's 3am here and I must wake up for my paying job tomorrow.

 

My Software solution makes it easy. That is the whole point of why I wrote it - to make my tests easy to understand. A short and sharp learning curve is required, but it is simple after that.

 

Providing your device puts data into the serial port (usb) in grams, then everything else will work perfectly.

[Baldor]

As long as it accepts text files is fine for me. My data acquisition computer runs Linux.

 

How will you calculate altitude? (And instant speed maybe?) Mi physics is very rusty, and my math mouldy.

[stix]

My software allows you to import basic text files providing the sample rate is given. This is good because then you can import data from other software or a simple data text file.

 

On the recording side, the software is strict, but not too hard to provide. As I said above, to "record" using my software, what is required is a stream of data delivered to the serial port. Easy. Done.

 

Altitude calculations require a lot more thought. I have done this before using various methods, ie. using other software, using a notepad, pen and a calculator. For sub-sonic rockets, I think I'm ok with the math. How I implement that into my current software is the hard part - but doable.

 

There are other people on this forum who have done this. Maybe worth your while looking into that as well.

 

The equations I've used before, and the Authors of it, have escaped me at this moment. Even google doesn't help.!! Not too hard, and will be included in the final version.

 

In any case, when I post the software, just use it for what it is. I'll bother with the altitude calculations if I think it's worth it. It is a work in progress. Like I said, there are other solutions posted here.

 

The implementation of "Fuel Loss Compensation" for vertical test stands is much more difficult than basic altitude calculations.... Indeed!!

Edited March 12, 2018 by stix

[stix]

As long as it accepts text files is fine for me. My data acquisition computer runs Linux.

 

How will you calculate altitude? (And instant speed maybe?) Mi physics is very rusty, and my math mouldy.

 

Oh poop!! I just realized that your data acquisition computer is running Linux. I can only build for Windows 7,8,10. Are you able to have a dual boot system?

 

The altitude calculations are from a book called "Handbook of Model Rocketry" by G. Harry Stine. The calculations look very thorough. I've used similar equations in the past but not sure if they are the same ones. It's been a while.

 

All going well, there will be a graph showing velocity, acceleration and altitude over time. The user will be able to select any time period in the graph and see the actual data (for what point, I'm not so sure) but there will be the usual final list of results.

 

---

 

This is similar to the type of load cell amplifier I have used:

https://www.ebay.com/itm/4-20MA-Load-Cell-sensor-Amplifier-Transmitter-strain-gauge-transducer-0-5V-0-10V/262675931462?hash=item3d28b4a146:g:Ec4AAOSwh2xYAjaW

 

I have an explanation of my test rig here (post #5):

https://www.amateurpyro.com/forums/topic/12717-creative-method-of-measuring-thrust/?hl=%2Bload+%2Bcell+%2Bamp&do=findComment&comment=175394

 

In addition I use another smaller circuit board to increase the sample rate. The maximum I can get is 250Hz because of the A/D conversion time. I wish the Arduino had a 16bit version. The Pic 16 sounds interesting - I've done a bit of Pic programming in the past.

 

In any case, I'm happy with my test rig at the moment. It would be good if someone could test out my software., ie. beta testing. My brother did some Alpha testing for funtionality via teamviewer, but I'm the only one that has used it with a load cell attached.

 

Anyway Baldor, this is your thread, and it should be about YOUR tests. Maybe pm me if you're interested in the software.

Edited March 14, 2018 by stix

[Baldor]

No problem Stix, I capture the data in a Linux system, but I process it in a windows 10 system.

 

I have seen the ADC you are using, and it's very capable. It have an integrated amplifier with programmable gain. Could be possible to connect the load cell to the ADC without an amplifier?

 

Of course I will beta test your software, but only the data analisis part, due to the limitations mentioned above.

[stix]

Ok, PM me and I'll send you a link via wetransfer.

 

[EDIT] PM an email address.

Edited March 14, 2018 by stix

[Baldor]

So, I made two changes at a time, something that should not be done while testing. I changed the load cell to an smaller on(20Kg), and the BP, 75/15/10 instead of 60/30/10.

 

And got this:

 

 

Wow... 84s specific impulse!!

 

So I tested another motor, 60/30/10 this time, and te results were consistent with previous tests.

 

To be really sure, I repeated the test with 75/15/10:

 

With the 20Kg load cell get similar results.

 

 

 

2018-03-18 _testA_20KG.txt

 

And with the 40Kg load cell I have been using before:

 

 

 

2018-03-18 _testB_40KG.txt

 

Notice the peak at about 1.7s

 

Seems some BP get stuck to the walls of the tube, and the bulkhead was very thin (About 1/3ID or less), and a little fire pased the bulkhead.

 

Now:

 

75/15/10 and no CATO? I think My BP is tame. I will try vine charcoal.

 

Mi process is:

 

Pre-dry the components in a toaster oven.

 

Ball mill for 4 hours.

 

wet a little and pass it trough a sieve.

 

Let it dry.

 

Some suggestion for a better process?

 

Meanwhile, I will finish the 19mm tooling and make the same battery of tests.

 

 

 

[Arthur]

If 75/15/10 doesn't cato then look at two details, 1/ the powder is really poor, 2/ the nozzle is too wide. I'd expect 75/15/10 to be fuel for a fast nozleless motor

 

Revisit your BP technique. 75/15/10 all dry before weighing and milled to completion. Use alder or willow wood for charcoal. With a nozzle there should be no need to go to fast exotic charcoals

[Baldor]

I'm inclined to think it's my BP. Commercial pine charcoal. I will try with vine charcoal, and also longer milling time. Reducing the nozzle to 5mm (1/3 ID) is another experiment I should try.

 

BTW, motors are pressed to 6000PSI with a bottle jack press.

[Arthur]

FIRST make sure that your BP is milled to either as fast as you want it, or as fast as it can be using the ingredients you have. Pine is fine for BP it's also a good wood for sparkly tails. Put some ingredients in the mill and set to mill as usual. Remove a sample every (say) hour, and measure the speed of burn and cleanness of burn. Usually powder increases in speed with milling time with reducing gain in the last hours til no further gain is achieved. I suspect that your powder is not as fast as BP can be. Most people make a rocket cato then just slow the powder down a little til the motor is as fast as can be just short of CATO. If you can press BP into a cored rocket form with a nozzle without a cato then you powder is slow. Maybe make a batch of powder with 1 hour more milling and a batch with 2 hours more milling then you use at the moment.

[stix]

Baldor, I don't understand why you think your fuel is no good. The design of your motors, as far as I can determine, uses your fuel effectively and efficiently.

 

If the performance results are correct then your "E class" motor is similar to "Estes" E class ( if that matters). Have a look at the attached table.

 

The "E" motors on that chart use around 36 grams of bp for a total impulse of 30N-sec. You're around that mark with 26N-sec for 30 or so grams of fuel. That, in my opinion is a very good result.!!! So you have reached a very good "standard".

 

The current configuration of your motors, ie. core/internal burner, is pushing the limit for "nozzled" motors. I don't understand what it is you want to achieve? If you use faster burning fuel, then you will get a faster burning rocket motor with more thrust but less burn time, with similar total and specific impulse but It will be like a bullet! There is a limit before CATO. Fast burning fuel is more suited to "nozzleless" OR "End Burners" which require smaller nozzles (ie. Estes). I think your fuel is performing well - you got 84s of specific impulse. That is a good measure from black powder.

 

I did a possible altitude calculation scenario using some old software I wrote back in 2011. I used your testA as the example, and added some extra weight using twin 4mm (sticks on each side) with a rounded nose. If the calculations can be believed, then your current design is... well, have a look at the results.

 

I'm not sure how accurate my calculations are, but I've checked against two online calculators and the results are "similar" but certainly not the same because it depends which equations you use. After-all, It IS "Rocket Science" You may want to pm Dagabu or Peret on this forum. They already have software that does similar calculations. It's always useful to substantiate data and results from other sources.

 

Cheers.

 

 

btw. Baldor, I forgot to mention that I really like your "test bunker". nice and safe to do testing - can I have one?

Edited March 20, 2018 by stix

[Baldor]

Just build a fire pit for barbecues... :-) Two layers of bricks, the first refractory, the second normal bricks :-) Since there are only some metal stairs in front, seems pretty safe.

 

I'm not unhappy with the motor performance. I'm unhappy with the BP performance. The current 75/15/10 BP should cato, and need some taming.

[stix]

Wow... 84s specific impulse!!

 

Yes, that is a good result, but an important thing to think about is that burning fuel over a shorter period (approx. 0.36s in your case) will always yield a higher specific impulse because it is much more efficient that way - ie. like a bullet.

 

Making "usable" rocket motors is a compromise, a balance between waste of fuel and a reasonable, slower "take-off" depending on the purpose.

 

My preference is for a slower take-off, then increasing in thrust. That way I can at least see it take off. Especially true with model rockets. Afterall, if you take the time to make the body and fins etc. then it seems like a waste of effort if it takes off like a bullet (you won't see much). More so with model rockets than pyrotechnic rocket motors that serve the purpose of delivering a shell. It does depend on what you are wanting to achieve.

Edited March 24, 2018 by stix

[Baldor]

I'm going to use them for pyrotechnic, but I think they are a little too fast burning.

 

Now I'm just validating all the process, and finding the weak spots in it. I found my BP is too tame, so now i'm working in characterising it and make it faster, since it is the same BP I'm going to use for breaking. Once found what is the hottest I can do, I can always slow it for rockets, or make a batch the "old" (actual) way specifically for rockets, since seem to work well enough.

[Baldor]

It's possible for a 15mm core burner to go supersonic? I just made a rocket with the last of my current batch of 75/15/10, added some wikonur 25 as delay for the effect, put a stick on it and launched.

 

It took off like a bat out of hell. A very short woooshh and then clac!!, like a whip. I almost lost sight of it. It just went up, and up, and up, with a short tail. When the tail extinguished, it was still going up.

 

Seems I have a decent base to start launching some shells.

[stix]

It's possible for a 15mm core burner to go supersonic?

 

I think it's possible.

 

This is a link to an online calculator for basic altitude calculations. You should give it a go and see what you come up with.

http://www.unm.edu/~tbeach/flashstuff/RocketAltitudeFixedSize.html

[dynomike1]

Given 2 identical motors which one will lift more? Nozzless or nozzle?

Edited March 26, 2018 by dynomike1

[stix]

So, you mean "identical" as in fuel weight, and external diameter, and perhaps length - Lift more, for what amount of time?

 

It's a hard one, but I'll put my money on the one with the nozzle would deliver more total impulse. The nozzleless may lift more as in greater initial thrust, but not higher in altitude. That's where the nozzle makes the difference. That's why "commercial" rockets have nozzles - efficiency.

 

Nozzleless motors are more about convenience than efficiency. Over the short term thrust, it may seem beneficial but at the expense of fuel weight - but if you are lifting a pyrotechnic shell then I can see the point.

 

Which design lifts more? I'll still go with the nozzled variety for lifting the same weight to a higher altitude, and the nozzleless for lifting more weight over a shorter distance.

 

I don't think there is a correct answer because we don't know the distance (you have given none) nor the weight required to lift.

 

[EDIT] In any case, and as interesting as it is Dyno, these sort of ideas and questions should be posted in a new thread. This thread is about Baldor's experiments and we should not interfere with that.

Edited March 26, 2018 by stix

[justvisiting]

I compared 2 rocket motors made with willow BP on 1lb BP tooling. One had a nozzle, and one didn't. Both were J-hooked with visco. Obviously the one with no nozzle had slightly more propellant. The nozzled motor had higher impulse, higher peak thrust, and would lift a rocket weighing a total of 1 1/2lbs to an altitude of 515 feet. The nozzleless motor would lift the same rocket to 'only' 435 feet. The nozzled motor had a burn time of .35 sec, and the nozzleless BT was .4 sec. These results are pretty typical for milled black powder on black powder tooling. The less the powder is milled, the greater the differences will be.

[Baldor]

Next test will be a nozzle vs nozzleless, Dino. Nozzleless will not be the ideal one since I will use the same tooling than for nozzled.

 

This is what a test stand is for. :-)

[dynomike1]

Thanks i got my answer.

[Baldor]

As promised, nozzle vs nozzleless. Both with the same tooling and same BP

 

Nozzled:

 

 

Worst results than the last test. A new batch of BP, very, very dry and milled for 10 hours. Maybe a little water in the BP improves rocket performance? Still, not bad.

 

Nozzleless:

 

 

Not bad at all. Less peak impulse, but almost same total impulse. I think I will go nozzleless from now on. Time to put some staninless in the lathe and make new tooling.