Open Rocket Launch Rod Length

The question of what Launch Rod Length to enter into Openrocket has come up in a couple of threads on TRF. People on TRF had said that in order to accurately simulate rocket stability as it clears the Lachlan rail you should subtract the height of the top rail button from the launch rail length. However I had noticed code within OR that appeared to be doing this calculation, however I was then told that testing seemed to contradict this. So after further tests myself and a bunch of digging into the code I can confirm that although OR does calculate the effective launch rod length during the simulation it does not actually use this value anywhere.

So the short answer is yes you should enter the the effective launch rod length in the simulation options not the actual length, but it’s not quite that simple...

The calculation in OR was a different then described above which had me initially confused as to why it was a coded this way. After some thought I realised there are 3 main scenarios and the code was only designed for 2 of them which may possibly be why it is not used. I was however able to modify the code to work as originally intended, but not sure if it should be used due to the missing functionality, hence this post to discuss the issue further.

So now to the details about what cases I think should be covered, I’ll go from simple to complex.
1. The first case is where a Launch Tower is used. If the rocket has no Launch Lug the the assumption in OR is that a Launch Tower is used an OR uses the actual Launch Rod Length.
2. The second case is where a Launch Lug is used. Note that it is important to distinguish between a Launch Lug and a Launch Button, in that a Launch Lug will not allow the rocket to pitch or yaw when only one Launch Lug is still on the rail. If the rocket has a Launch Lug then OR is that calculates the Effective Launch Rail Length as the actual Launch Rod Length minus the height of the lower Launch Lug.
3. The last case is where a Launch Button is used. As previously mentioned a Launch Button will allow the rocket to pitch or yaw when only one Launch Lug is still on the rail. OR does not have any provision for indicating this type of hardware so it’s not as simple as adding an additional calculation. In addition the behaviour of the rocket with one button on the rail is complicated.

I’ll post more details about case 3 later as I need to go out.

Look forward to this discussion.

Good investigation of the code.

I think one button in the rail would not roll nor pitch. Yaw only?

Buckeye said:

Good investigation of the code.

I think one button in the rail would not roll nor pitch. Yaw only?

Click to expand...

Roll I would expect a tiny bit due to slight differences in the fit between the button and rail but may get worse due to wear, I was going to ignore this but perhaps we should look at it too. Pitch is where rail strike comes into play, definitely limited in angle by the body of the rocket but does obviously happen. Perhaps some people with access to a rail at home can fit their rockets and come up with some maximum angels for all three axises? This will give us a ball park about the limits expected. If you don’t have anything to measure the angle take a pic and do the trigonometry.

So some more details about the complexities of case 3 - that is what happens to the rocket while one rail button is still on the rail. As perviously discussed with one rail button the rocket becomes partly free to rotate, particularly yaw, but also some pitch and perhaps very limited roll.

To explain why this is an issue we need to go back to basics about rocket stability and how it works. So with a rocket in free flight, if the rocket does not have a zero angle of attack to the wind then it will experience a force at the Center of Pressure, this results in the rocket rotating about the Center of Gravity and correcting the AOA. Which is of course why the relationship between these points is so important to stability.

Now let’s look at what happens while the last rail button is still on the rail. The difference is that the rocket is no longer free to rotate about the CG instead it is restrained by the rail button so will rotate at that point (albeit with some limitations). What is particularly bad is that typically the last rail button is located at the far aft of the rocket well below the CP so it will rotate away from the angle of attack. IE the rocket has a big negative stabibily margin.



So from a design point having a low rail button is not a good idea. It would be better to use a Launch Lug for the lower mount point or move the Rail Button up as far as practical towards the CG or at least 1 caliber of stability, infact any improvement on the negative stability would be helpful especially in windy conditions. Again I’m talking about typical design, one less common use of rail buttons is on fly away rails, as these sit above the fins they will definitely perform better especially as the distance between the buttons are short so less time for a deviation in flight to occur.

So the next question is how should this be handled? Until we run some numbers I’m not sure about how much an issue this is, although we do know from rail strike that it is happening. Also there’s a question about how to integrate this scenario into OR as it does not have an option for it and it effectively creates an alteritive flight condition that needs to be simulated. Also there might be issues around the moment of inertia, have to figure that out and check what the code is doing.

Interested to hear peoples thoughts on this and possible work arounds.

Meh, I've always just simply put in effective rod length. My club has 6 and 8 foot rails. The blast plate is 6 inches from the end, from there I subtract the location of the TOP rail button to get effective length for my rockets. Might be tedious, but it works well.

Since we've been working OR for how long now WITHOUT it doing the computation for us, and getting pretty darn good ballpark numbers, I'm not really sure that we have a 'problem' more than an unexpected non-implementation of an equation that OR computes.

Now if the OR code were to change to REQUIRE that you enter launch lug/rail button location to do a successful launch because of the computations, that would be a step backward IMHO, since simply leaving them off now only has marginal impact on CD and speed/altitude numbers.

A few thoughts I'll toss-in just to mess things up:

a) There are also rail guides such as this (Acme Conformal Launch Guide), which I guess I would have considered a 'button' since it uses the same rail that a button does, but it would act more like your #2 than #3 given its shape. These won't rotate (any more than a lug would) since they're rectangular instead of round.
b) I'll admit since I've been a BAR I've only flown HPR and only used 1010/1515 rails (with buttons or the conformal guide I just described, or fly-away which is next), but it seems to me that there's a suggestion above that #2 is more 'stable' (perhaps solid is a better word, not allowing any pitch/yaw) than #3, but that seems to ignore that the rods themselves can be much less sturdy than rails, so I think (certainly for rockets with a larger wind profile) if a rocket catches wind while using lugs on a rod it's more likely to just bend the rod as it nears its end (which may cause the lug(s) to try to rotate the entire rocket), where something like a 1010 rail isn't likely to bend along its length, not that you might still get tilting of the pad or the like in more extreme cases even with a rail. So just noting that while a single button could allow a rocket to rotate on a rail, it's not like a lug on a rod is going to be perfectly rigid (and the rod can bend more than a rail even while both lugs are still engaged), so it seems pointless to consider one case without also considering the other (which in my mind suggests neither one is really worth worrying about, rod whip would probably be an even more complicated issue to try to predict/handle).
c) In my mind it would be more valuable to consider fly-away lugs/buttons, as something that would not affect the aerodynamics (assuming OR's current lugs have an affect on drag, maybe they don't) once in flight, but would still be useful for the "on the rail" calculations as long as the fly-away device is still in contact with the rod/rail. It also does have an affect on the rocket early in the launch, with no way in OR to say that it is removed from the rocket very shortly after the rail is cleared. So something that could indicate the parameters of the fly-away guide (extra weight, namely, presumably the aerodynamics don't matter much since the rocket won't be going too fast before they should detach) and consider rod/rail attach up to the point where it clears the rod/rail would be useful in my opinion.

This is an important discussion. Here are a few more points to consider.

• 1010 rails *will* bend. At 8' long with a 10lb rocket it gets pretty noticeable
• A single launch lug usually will allow a fair amount of pitch and yaw deviation, depending on the lug length, material (cardboard deforms), and how close-fitting it is.
• Binding torque is a serious consideration. If there is any noticeable amount, the launcher exit velocity can be a lot less than what the simulation says.
• High thrust motors deliver a lot of force into the blast deflector. This can easily be enough to cause deflection in the pad legs and change the launch angle significantly.

FYI I forwarded this thread to the OR dev mailing list. Folks are watching.

I agree that fly-away guides would be a useful feature.

Banzai88 said:

Meh, I've always just simply put in effective rod length. My club has 6 and 8 foot rails. The blast plate is 6 inches from the end, from there I subtract the location of the TOP rail button to get effective length for my rockets. Might be tedious, but it works well.

Since we've been working OR for how long now WITHOUT it doing the computation for us, and getting pretty darn good ballpark numbers, I'm not really sure that we have a 'problem' more than an unexpected non-implementation of an equation that OR computes.

Now if the OR code were to change to REQUIRE that you enter launch lug/rail button location to do a successful launch because of the computations, that would be a step backward IMHO, since simply leaving them off now only has marginal impact on CD and speed/altitude numbers.

Click to expand...

No chang in that regard, if you leave them off the behaviour would be identical.

woferry said:

A few thoughts I'll toss-in just to mess things up:

a) There are also rail guides such as this (Acme Conformal Launch Guide), which I guess I would have considered a 'button' since it uses the same rail that a button does, but it would act more like your #2 than #3 given its shape. These won't rotate (any more than a lug would) since they're rectangular instead of round.

Click to expand...

I would consider these as Launch Lugs they are definitely not a Button.

woferry said:

b) I'll admit since I've been a BAR I've only flown HPR and only used 1010/1515 rails (with buttons or the conformal guide I just described, or fly-away which is next), but it seems to me that there's a suggestion above that #2 is more 'stable' (perhaps solid is a better word, not allowing any pitch/yaw) than #3, but that seems to ignore that the rods themselves can be much less sturdy than rails, so I think (certainly for rockets with a larger wind profile) if a rocket catches wind while using lugs on a rod it's more likely to just bend the rod as it nears its end (which may cause the lug(s) to try to rotate the entire rocket), where something like a 1010 rail isn't likely to bend along its length, not that you might still get tilting of the pad or the like in more extreme cases even with a rail. So just noting that while a single button could allow a rocket to rotate on a rail, it's not like a lug on a rod is going to be perfectly rigid (and the rod can bend more than a rail even while both lugs are still engaged), so it seems pointless to consider one case without also considering the other (which in my mind suggests neither one is really worth worrying about, rod whip would probably be an even more complicated issue to try to predict/handle).

Click to expand...

Ceartainly modeling material properties such as flex of rods etc is way beyond what OR is currently doing and requires a lot of details of those material properties akin to what is required to use Fin Sim. As for modeling a rail being titled, yeah it’s pretty extreme and very rare occurrence and again requires addional detail about the design of the rails / stand and even how they may have been setup on the day.
The whole discussion came about as people want to know “is my rocket stable leaving the launch rod/rail” and then questions about “what the effective rail Length is” so in that regard there are some easy improvements that can be made. As for my further question about how the rocket rotates while still on the rail, I’d like to get some sort of numbers on this before dismissing it as not worth while worrying about. Unlike flex / rail being titled a reasonable amount of the functionality is there it just needs to be tweaked for this scenario.

woferry said:

c) In my mind it would be more valuable to consider fly-away lugs/buttons, as something that would not affect the aerodynamics (assuming OR's current lugs have an affect on drag, maybe they don't) once in flight, but would still be useful for the "on the rail" calculations as long as the fly-away device is still in contact with the rod/rail. It also does have an affect on the rocket early in the launch, with no way in OR to say that it is removed from the rocket very shortly after the rail is cleared. So something that could indicate the parameters of the fly-away guide (extra weight, namely, presumably the aerodynamics don't matter much since the rocket won't be going too fast before they should detach) and consider rod/rail attach up to the point where it clears the rod/rail would be useful in my opinion.

Click to expand...

I’ve looked at this sort of issue before and made alternative models specifically for looking at how fly away rails effect stability and altitude https://forum.ausrocketry.com/viewtopic.php?f=6&t=5312&hilit=Purple+parrot&start=60#p62327. Keep in mind that I was not aware of issues with effective rail length at the time and I later came to the conclusion that I should have held onto the fly away rails for much longer. I think we could definitely get some functionality into OR for fly away rails without too much an issue, but probably not as comprehensive as what I was aiming for in my Purple Parrot build (at least at this stage).

caveduck said:

This is an important discussion. Here are a few more points to consider.

• 1010 rails *will* bend. At 8' long with a 10lb rocket it gets pretty noticeable

Click to expand...

Please see above response about launch rod flex and the difficulties of modeling material properties.

caveduck said:

• A single launch lug usually will allow a fair amount of pitch and yaw deviation, depending on the lug length, material (cardboard deforms), and how close-fitting it is.

Click to expand...

Are you able to please get some details on this? I’m interested in actual numbers, I’m thinking there could be an option to enter limits of movement into the OR model. With regards to material flex again this is way beyond what OR is capable of...
Also I’ve realised that Launch Lug on a rod allows full 360 degree roll.

caveduck said:

• Binding torque is a serious consideration. If there is any noticeable amount, the launcher exit velocity can be a lot less than what the simulation says.

Click to expand...

True, but extremely hard to simulate. There’s a whole bunch of variables such as materials, dirt, rust, lubricant etc. What we could possibly do is add a warning if the rocket rotates to the max allowed angle. This is when binding or rail stike is going to occur.

caveduck said:

• High thrust motors deliver a lot of force into the blast deflector. This can easily be enough to cause deflection in the pad legs and change the launch angle significantly.

Click to expand...

Again would require a whole bunch of details about the launch rail/stand design. Possible I guess, but outside the scope of what we are looking at here. I don’t think people are going to go to the bother of getting these details and entering them unless they are looking at stand design for the most part.

I guess what I am looking for here was to first get the effective rod/rail length calculations working as intended.
Secondly I wanted to get more accurate details on what was happening on high altitude flights as the angle that rocket leaves the pad affects the altitude and recovery area and a few degrees can make a difference. My assumption here is of course a properly designed launch pad etc.
Thirdly I’m of course concearned about marginal launch conditions and getting better simulations for these.

neil_w said:

FYI I forwarded this thread to the OR dev mailing list. Folks are watching.

I agree that fly-away guides would be a useful feature.

Click to expand...

Saw that, thanks.

Thought I should add some pictures of various lugs / Button so we know what we’re all talking about and can characterise them.

Launch Lugs for rods - Allow 360 degree roll.



Linear Rail Lugs - Allow very limited rotation in any axis.



Rail Button - Allows yaw and some pitch, very limited roll.



Airfoil Shaped Rail Button - These are really a cross between a button and a Lug, they allow limited rotation.



Acme Conformal Launch Guides (as mentioned previously)
Top: Launch Lug for rod - Allow 360 degree roll.
Bottom: Linear Rail Lug - Allow very limited rotation in any axis.




Fly Away Rails



Please post any other that you feel are substantially different to these.

Though I mentioned the items I did more in the context of un-modeled effects that can possibly overwhelm the modeled ones, the deflection load of 8020 rail is pretty easy to characterize since they provide the yield and modulus for their products, and the beam deflection calcs are straightforward. Data for round rods of various materials is readily available. 8020 themselves have a deflection calculator for all of their rails at https://8020.net/deflection-calculator . A 1 lb lateral load at the end of an 8 foot 1010 rail generates a deflection of 3 inches, which is substantial. Using 1515 rail cuts it to 1.2 inches.

The possible pitch and yaw deflections allowed by a rod thru a single tubular lug are geometrically computable from dimensions only, if you assume the lug is perfectly rigid (which is very false in cases not involving metal lugs).

Binding forces on a lug would be pretty hard to simulate since it will be material and geometry and rod/rail contamination dependent.

Agree that launch stand deflection would have to be empirically tested. I'm mostly bringing it up b/c I've seen crashes and broken launchers from that cause; people should not think that just because OR shows a reasonable exit angle from the launcher that it will necessarily be true!

Personally I don't think any of these features *have* to go into OR. From a computation standpoint, it *could* calculate rod/rail deflection arising from crosswind loads at launch, and perhaps from motor thrust axis offset, and issue a warning if the deflection exceeded some threshold.

Wow, 3 inch deflection with only 1# of force? That's amazing... apparently these things aren't as stiff as they seem.

caveduck said:

Though I mentioned the items I did more in the context of un-modeled effects that can possibly overwhelm the modeled ones, the deflection load of 8020 rail is pretty easy to characterize since they provide the yield and modulus for their products, and the beam deflection calcs are straightforward. Data for round rods of various materials is readily available. 8020 themselves have a deflection calculator for all of their rails at https://8020.net/deflection-calculator . A 1 lb lateral load at the end of an 8 foot 1010 rail generates a deflection of 3 inches, which is substantial. Using 1515 rail cuts it to 1.2 inches.

Click to expand...

Those calculations are for a static load where as a Launch is extremely dynamic. Basically the rail is going to viabrate complicating the deflection. FYI the average angle for a 3 inch deflection on an 8 foot rail is about 1.8 degrees.

caveduck said:

The possible pitch and yaw deflections allowed by a rod thru a single tubular lug are geometrically computable from dimensions only, if you assume the lug is perfectly rigid (which is very false in cases not involving metal lugs).

Click to expand...

Yes could be done, it does require that these dimensions along with the rail dimensions to be entered. Also complicated when you need to consider where striking the body or fins occurs etc. The other consideration is wear on the rail / button.

Banzai88 said:

, since simply leaving them off now only has marginal impact on CD and speed/altitude numbers.

Click to expand...

Not always true. In OR, put simple launch lugs on a small diameter rocket and watch the drag and altitude change 10%-20%. I manipulate the lug dimensions to "look like" a rail button, but I am not sure that OR calculates this correctly. I wish OR had a dedicated rail button drag component like RASAero.

Anyway, I think the purpose of this discussion is to use the launch guides in the simulation for better prediction of the rail departure event.

Another thought on something that can impact the effective launch rail length, whether the rocket sitting upright on the rail rests on the bottom rail button (i.e. the stop is only deep-enough to interfere with the button but not the airframe itself) or if the rail stop rests on the bottom of the rocket (which could be airframe or motor nozzle). I've had all 3 with my local launch clubs, sometimes at the same launch due to differences from pad-to-pad. If the lower button is close to the bottom of the rocket it might only make an inch or two of difference, but in other cases (like the fly-away guides that sit above the wings) it can be more significant. Plus some motor nozzles stick out quite a bit more than others (probably another inch or maybe two for the 38/54mm motors I've used), not sure if OR's motor data even has that detail.

woferry said:

Another thought on something that can impact the effective launch rail length, whether the rocket sitting upright on the rail rests on the bottom rail button (i.e. the stop is only deep-enough to interfere with the button but not the airframe itself) or if the rail stop rests on the bottom of the rocket (which could be airframe or motor nozzle). I've had all 3 with my local launch clubs, sometimes at the same launch due to differences from pad-to-pad. If the lower button is close to the bottom of the rocket it might only make an inch or two of difference, but in other cases (like the fly-away guides that sit above the wings) it can be more significant. Plus some motor nozzles stick out quite a bit more than others (probably another inch or maybe two for the 38/54mm motors I've used), not sure if OR's motor data even has that detail.

Click to expand...

That’s a good point, the calculations as they stand assume that the end of the rocket lines up with the bottom of the rod / rail, so the height of the Lug from the end of the rocket is subtracted from the rod / rail length. This is the more conservative value so should atleast be safe if the rocket does sit down further. I think the motor itself will be considered when the bottom of the rocket is obtained, need to confirm that I guess.