My Center of Pressure Test Models

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Jeff Lassahn

Well-Known Member
TRF Supporter
Joined
Jul 5, 2020
Messages
552
Reaction score
779
Location
Portland OR
Following the tradition of test-driven development, I've been making a collection of OpenRocket models to use as test cases for trying out center of pressure and stability calculation algorithms. Some of these are real rocket designs, some of them are simple geometric shapes that have interesting properties, and some of them are specific attempts to examine interesting corner cases.

Now that I've got a collection, I figure I'd start posting them here in case anyone else is interested.

My interest here is only in understanding stability estimation, so although a few of these are flyable designs many of them don't have proper materials, wall thicknesses, masses, internal structures, etc assigned. In many cases the only thing that matters for my purposes is the surface shape. So don't try to interpret things like the Center of Gravity in most of these.

I'm also not currently thinking about drag coefficient estimation or altitude determination, only stability, and I'm also mostly focused on sub-sonic low Mach number behavior.

So here's a few to start...
 
"The Original"

This is a rough model of the Aerobee 350. This design was used by James Barrowman in his original thesis where he laid out the first version of the "Barrowman equations". He uses it as the main experimental test for his system, comparing against a CoP measured by wind tunnel tests.

As one might expect, the Barrowman equations and OpenRocket do a great job of CoP estimation for this design.
Screen Shot 2024-03-11 at 10.46.07 PM.png
 

Attachments

  • Aerobee350.ork
    1 KB · Views: 0
"Consider a spherical rocket"

This is a sphere. Probably the airflow around this shape is more studied than any other realistic geometry. Because if its symmetry reliable solutions are known for a wide range of conditions, and lots of test data has been collected.

The Center of Pressure of a sphere is in the center. Everyone's pretty sure about this.

Sadly, OpenRocket gets this one very wrong, estimating the CoP at negative infinity. Mostly because of problems with handling the trailing edge of blunt bodies, which will be explored in several later examples.
Screen Shot 2024-03-11 at 10.47.14 PM.png
 

Attachments

  • Sphere.ork
    710 bytes · Views: 0
"It's Flat"

This one is a circular disk face-on to the airflow.
It's interesting both because it's another simple, well-studied geometry, and because Bruce Levison uses it as a starting point for his base-drag related stability corrections (the "Base Drag Hack").

It's an example where whether the rear of the model is left open or closed off makes a huge difference in the results of Barrowman-like calculations. If you leave it open, they estimate the CoP at the center of the disk. If you close it, they estimate the CoP at negative infinity.

The effective CoP is actually substantially behind the disk, making this a great example of a case where the Center of Pressure is in empty space outside of the object.

Screen Shot 2024-03-11 at 10.47.45 PM.png
 

Attachments

  • Disk_open.ork
    977 bytes · Views: 0
  • Disk_closed.ork
    979 bytes · Views: 0
"It deserves a catchy name because it's cool looking"

The classic streamlined teardrop shape. This is a traditional low-drag shape, and I'm surprised that searching the web doesn't turn up very much analysis about it. Everybody loves this as an airfoil cross-section but not so many people seem to research it as a rotationally symmetric body shape.

I wish I had better data about the actual CoP of this shape, my best guess is that it should be about 1/4 of the length back much like is true of airfoils.

Screen Shot 2024-03-11 at 10.48.19 PM.png
 

Attachments

  • Streamline.ork
    686 bytes · Views: 0
"Not much call for it? It's the most popular rocket geometry in the world!"

This is a model of the Congreve 32 pound artillery rocket. It's a classic of the history of rocketry, and also an example of a class of rocket shapes we've all seen lots of places as fireworks, etc.

It's also really weird from a stability modeling standpoint, and not well handled by existing tools.

Screen Shot 2024-03-11 at 11.24.00 PM.png
 

Attachments

  • Congreve.ork
    870 bytes · Views: 0
Following the tradition of test-driven development, I've been making a collection of OpenRocket models to use as test cases for trying out center of pressure and stability calculation algorithms. Some of these are real rocket designs, some of them are simple geometric shapes that have interesting properties, and some of them are specific attempts to examine interesting corner cases.

Now that I've got a collection, I figure I'd start posting them here in case anyone else is interested.

My interest here is only in understanding stability estimation, so although a few of these are flyable designs many of them don't have proper materials, wall thicknesses, masses, internal structures, etc assigned. In many cases the only thing that matters for my purposes is the surface shape. So don't try to interpret things like the Center of Gravity in most of these.

I'm also not currently thinking about drag coefficient estimation or altitude determination, only stability, and I'm also mostly focused on sub-sonic low Mach number behavior.

So here's a few to start...

Nice collection for study. However, didn't you mention in the CoP with CFD thread that we should not get hung up on center of pressure abstractions? :D

Spheres, plates, bottle rockets, etc. are definitely abstractions in the sense of Barrowman and OpenRocket.
 
"It's Flat"

This one is a circular disk face-on to the airflow.
It's interesting both because it's another simple, well-studied geometry, and because Bruce Levison uses it as a starting point for his base-drag related stability corrections (the "Base Drag Hack").

It's an example where whether the rear of the model is left open or closed off makes a huge difference in the results of Barrowman-like calculations. If you leave it open, they estimate the CoP at the center of the disk. If you close it, they estimate the CoP at negative infinity.

The effective CoP is actually substantially behind the disk, making this a great example of a case where the Center of Pressure is in empty space outside of the object.

View attachment 635042

Got any references for these studies showing the effective CP? Levison did not provide any, and I could not find one with casual Googling.
 
Got any references for these studies showing the effective CP? Levison did not provide any, and I could not find one with casual Googling.
Unfortunately, no. One of the problems that seems endemic in the model rocket literature is that when papers get published they appear without references that let someone backtrack to their original data or sources.

About the disk CP specifically, I don't have any published data to quote.

Several months ago, I got annoyed enough about this that I did some experiments where I cut a disk out of cardstock, glued a stick onto the back, used modeling clay to adjust the CG to various points, and dropped it disk down from the top of some stairs. Not a particularly elegant or exact measurement technique, but I can confirm that the CP is about -2 diameters. plus or minus maybe 30%.
 
I'll be adding more of my own test cases in a few days, but for now I want to call attention to this thing I found while looking for other stuff:

https://ntrs.nasa.gov/citations/19690009748

NASA Technical Memorandum X-53770, which is a summary of some very thorough wind tunnel testing of Saturn V models. It reports CP (and some other aerodynamic characteristics) across 6 different design configurations, including with and without fins. There's high resolution plots for each configuration across varying angle of attach an Mach numbers both subsonic and supersonic.

TL;DR: the CP of the Saturn V Apollo full stack is about 3.5 calibers from the back, which is just a little behind the top of Stage 1. It moves around some as the vehicle goes trans-sonic.

I'll dig around and try to find my OpenRocket models to compare, but my memory is that they were a little too conservative, maybe 1 caliber farther forward from what the wind tunnel data shows.
 
The fascinating world of tail cones...

It's widely believed that tail cones are destabilizing. I can't help thinking that OpenRocket overestimates the effect, though. So there's this pair of test cases...

Unlike some of the others, these are basically buildable, and could be swing tested or flown to confirm real world behavior. My suspicion is that the no cone version's CP is pretty close to accurate, but that the cone version is not actually unstable as indicated.

Screen Shot 2024-03-23 at 4.47.53 PM.pngScreen Shot 2024-03-23 at 4.47.30 PM.png
 

Attachments

  • TailConeOn.ork
    2.1 KB · Views: 0
  • TailConeOff.ork
    2.1 KB · Views: 0
"They're kind of like minimalist Goonybirds"

So, how do short rockets with large body diameters really behave?
I'm thinking when the weather is better and I have some spare time, I'll put a nose on various lengths of BT-60 tubing, put a threaded rod with some nuts and washers in it so I can adjust the center of gravity to wherever I want it, and do a bunch of swing tests until I get a decent estimate of where the center of pressure is on several lengths of finless bodies.

I'm not exactly sure where the resulting CP will be, but I'm pretty sure it should be different for the different lengths and all of them should be farther back than shown here.

Screen Shot 2024-03-23 at 9.09.19 PM.pngScreen Shot 2024-03-23 at 9.08.49 PM.pngScreen Shot 2024-03-23 at 9.08.16 PM.png
 

Attachments

  • FinlessShort.ork
    916 bytes · Views: 0
  • FinlessMedium.ork
    916 bytes · Views: 0
  • FinlessLong.ork
    919 bytes · Views: 0
And one more from the swingin' '60s...

The Apollo capsule was stable when reentering the atmosphere heat-shield first, over a huge range of Mach numbers. It's got to be in trim starting at high hypersonic speeds and stay that way down to moderate subsonic speeds.

So here's a 1:100 scale model. The actual CP is somewhere between +4cm and +4.7cm.

Screen Shot 2024-03-23 at 9.22.15 PM.png
 

Attachments

  • ApolloCapsule.ork
    877 bytes · Views: 0
So. That's my test set for the moment. Some of them have known stability characteristics and are ready to compare against algorithms. Some of them are not so well understood (at least by me) and need some experimental measurements.

This set is very much designed to explore the stability characteristics of rocket bodies. It's easy to imagine other sets to explore other things, like fin shape, or interference between groups of components, etc.

More news if I make any progress on actually testing any of these...
 
"They're kind of like minimalist Goonybirds"

So, how do short rockets with large body diameters really behave?
I'm thinking when the weather is better and I have some spare time, I'll put a nose on various lengths of BT-60 tubing, put a threaded rod with some nuts and washers in it so I can adjust the center of gravity to wherever I want it, and do a bunch of swing tests until I get a decent estimate of where the center of pressure is on several lengths of finless bodies.

I thought about doing the same, but with flight tests. I designed a variable-CG rocket and got a bulk pack of motors. Maybe I will get around to flying it. For now, I found it easier and more reliable to use CFD.
 
Nice list of rocksim files. Please keep it going. I can't wait to see your results.
 
This is a pair of models that are trying to answer a very specific question I have about the pressure distribution over a flat disk.

It's common to claim that spool rockets and saucer rockets, etc. are stable because of drag forces on the rear surface of the disks. I suspect that this is not true, that the stabilizing force mostly comes from the Munk moment torque on the front surface.

So...
Screen Shot 2024-06-14 at 9.48.03 PM.pngScreen Shot 2024-06-14 at 9.47.36 PM.png
As always, the nose of the rocket is on the left. If the stabilizing force on a disk comes from drag on the rear, the top version will be more stable. If the force comes from the pressure distribution on the front, the bottom version will be more stable.

My personal guess is that the top one will have a CP around 10cm, and the bottom one will have a CP around 20cm.

Should be checkable with a swing test...
 
If you're going to mount something on a threaded rod and swing it, why not just hold the rod up in the air on a windy day? You could do it from a car, but it would probably need to be a long rod to avoid the effects of the car on airflow.
 
Should be checkable with a swing test...
Swing tests would not be definitive because of the lack of control over the AOA of the article. CP does vary with AOA, especially with long body tubes. That is how my Bellyfloppers work.

If you're going to mount something on a threaded rod and swing it, why not just hold the rod up in the air on a windy day? You could do it from a car, but it would probably need to be a long rod to avoid the effects of the car on airflow.

A simple "wind tunnel" made from a fan and flow straightener should work for small test articles.
 
Back
Top