Finding Reentry Aerodynamics Characteristics

In summary: I haven't found them being used for drag yet.Right now, the correction factors are used mostly for lift in ballistics.
  • #1
aero5682
18
0
This is a question about reentry aerodynamics
How are aerodynamic characteristics found? I know that aerodynamic equations are not that accurate, and the best way to do it is by simulation. I also heard that if you want it to be precise, you add corrective factors to the calculations. How do you work out aerodynamic characteristics to make predictions? Do you start with a simulation or whatever you use and then test it, and then add corrective factors to the simulation from the testing, or do you start with testing to work out things like ballistic coefficient and then put that into the simulation, or some other way?
 
Engineering news on Phys.org
  • #2
What "aerodynamic equations" do you mean when you say they aren't very accurate? Also, what exactly do you mean by "aerodynamic characteristics"? A little more precision in your language will net you a better answer.
 
  • #3
Basically, what is the process for finding aerodynamic characteristics and making a prediction? Mostly I am wondering about the corrective factors. For example, if NASA is predicting where a spaceship will land, do they always use corrective factors? Are the corrective factors different for each ship? By aerodynamic characteristics I mean ballistic coefficient, coefficient of drag and things like that that you would need to predict a landing site. The equations I mean are whatever you choose to use to predict a landing site. I am talking about predicting a landing site. Theres a lot of information, but its all pretty general about the actual process of doing it. You can run simulations and you can do expirements, but how does the info from the expirements effect the simulations and the predictions? Also, I am talking about a ballistic entry
 
  • #4
Generally you'd start with some modeling, then perform experiments in wind tunnels to try and validate the simulations, rinse and repeat until the answer is good enough. Drag is an incredibly hard thing to predict on a computer, and it is nearly impossible to match all characteristics of a re-entry-type flow in a ground test, so there is some iteration and guess work. Many quantities like drag and surface heat transfer also rely pretty heavily on models incorporated in the simulations based on theory.
 
  • #5
Is the purpose of the wind tunnel expirements to try your models and see what really happens, and to see how accurate the models are? How do the correction factors fit into the process? Are they always used, something that you can use sometimes, or rarely used? I've heard that theyre found expirementally, and they make the models fit what actually happens, is that right? I've seen them used in ballistics, and I've found NASA papers talking about using them for lift, but I haven't found them being used for drag yet. I am still trying to find that
 
  • #6
Without more information, I don't know what "correction factors" you are referencing.

The best I can tell you based on what you've said so far is that lift is comparatively easy to compute compared to drag or heat transfer since you can get a very close approximation of lift based solely on inviscid flow. That isn't true for drag and heat transfer, so you need to know a lot of information about the boundary layer, which is very difficult to handle computationally. As a result, researchers and designers typically rely on computational models based on various theories that average out the finer physics but still aim to get the major characteristics like drag correct. That makes the computational problem easier to solve, but it also requires verification and validation against experiment to make sure the various models are tuned properly, so it's usually an iterative process.

However, there are dozens if not hundreds of different models that can be used for the various phenomena, and the "correction factors" could mean anything depending on the model. I, myself, am not in the business of turbulence modeling or predicting the landing point of a re-entry vehicle, so I am not sure how much more I can help you there on specifics. I do experiments on the more fundamental aspects of high speed boundary layers.
 
  • #7
aero5682 said:
I know that aerodynamic equations are not that accurate, and the best way to do it is by simulation.
This makes no sense: if the equations didn't work, how could they be used to provide accurate simulations?
 
  • #8
russ_watters said:
This makes no sense: if the equations didn't work, how could they be used to provide accurate simulations?

I'm not so sure he actually meant it this way, but generally, most simulations are performed with sets of equations that have been simplified and aren't necessarily very accurate in general. They only work well for specific parts of the problem. That's sort of the basis of turbulence modeling.
 
  • #9
Ive seen the most about corrective factors in ballistics, where for example for artillery at a long range, theyll take the models and predict where it will land, then try it expirementally and see where it actually lands. Then theyll use the info from the expirement to make corrective factors to add to the prediction somewhere so its closer to where it actually lands. They don't know why the prediction is off, but the correction factors make it closer to where it actually lands. They leave working out why its off until there are better models. With reentry what I am thinking of might be what you were saying about tuning the models. How exactly do you do that? What I am thinking of is let's say you work out your models and go to test them and the drag is low. Mabey you add in a correction factor that's a number or mabey a variable based on some other things so your model says what it really is. Also about the accuracy, yeah I think that's what I meant, and more technical. Accuracy depends on how accurate you want it to be. Its extremely complicated calculations and simulations. Everything I've seen says the equations and simulations are closeish
 
  • #11
Modern aerodynamic simulations often utilize a turbulence model to try to model the turbulent drag and heat transfer. For example, one particular popular one is the k-ε model. Basically, the full equations for fluid flow are very complex and can't be solved in a usefully-short period of time, so you average out some of the problematic terms and model them with easier terms multiplied by a constant. Click the link I provided for a bit more detail.

Anyway, you can think of those simulations as an input-output machine and those constants as sort of like a set of dials you turn to adjust the output of the machine. Basically, you would make an educated guess based on physical intuition on what those dials should read and then check your output against reality, in this case, an experiment (or series thereof). Then you go back and tweak the dials and get an output that looks more like the experiments. Eventually, you have reasonable confidence and call it good enough. It is generally plenty accurate for predicting where something will land on a ballistic re-entry. Where it really turns into a crapshoot is in predicting the heat transfer rate into the surface, which is how you end up with bulky, overdesigned thermal protection systems.
 
  • #12
Needed some time to look into that. Really informative. Looks like the constants are the correction factors and where all the matching to expirements and tuning happens. Heres a NASA paper I saw about using correction factors for lift. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19760016071.pdf. It talks about adding correction factors based on expirements to make the models match the expirements. Is this an older method, or mabey what you do when your model just won't match to expirements? Or with all the different models you can use, mabey some require you to add in correction factors like this? It seems like you wouldn't need to add in correction factors when you have the constants you can tune. Also, I think the new capsule landed less than a mile from the target
 
Last edited:
  • #13
aero5682 said:
Also, I think the new capsule landed less than a mile from the target
That's a low Earth orbit. The Apollo missions were returning from the moon, much greater distance and speed involved in getting the reentry precise enough to end up within 5 km of target spot.
 
  • #14
Applying some sort of correction factor matrix to ##C_p## or downwash as discussed in the NASA paper is the same general idea as tuning a turbulence model to match experimental data. You adjust the correction factors until the final result you wish matches experiment, then you try to use the corrected theory to predict other situations.
 
  • #15
Another related thing, how do they correct for density differences from the atmosphere model theyre using? I've seen things that say it can be off by as much as 30%, which is huge
rcgldr said:
That's a low Earth orbit. The Apollo missions were returning from the moon, much greater distance and speed involved in getting the reentry precise enough to end up within 5 km of target spot.
Yeah that's true
 
  • #16
aero5682 said:
Another related thing, how do they correct for density differences from the atmosphere model theyre using? I've seen things that say it can be off by as much as 30%, which is huge
Sounds like you would need a control guidance system to compensate for differences from the standard atmosphere model like that.
 
  • #17
FactChecker said:
Sounds like you would need a control guidance system to compensate for differences from the standard atmosphere model like that.
Makes sense
 
  • #18
aero5682 said:
Another related thing, how do they correct for density differences from the atmosphere model theyre using? I've seen things that say it can be off by as much as 30%, which is huge.
I haven't found a suitable article about this, but if I recall correctly, the main issue is the altitude of transitions in atmosphere (each area between transitions essentially requires a different model, perhaps just different constants). Some the 1960's models have 11 transitions. The main discrepancies in terms of absolute (versus relative) altitude are for the outer transitions, where the atmosphere is thin. What I don't recall or know is how much difference in drag or heat variation this causes, but as posted by boneh3ad, they compensated with extra margin in heat handling, and attitude control (such as small thrusters).
 
  • #19
rcgldr said:
I haven't found a suitable article about this, but if I recall correctly, the main issue is the altitude of transitions in atmosphere (each area between transitions essentially requires a different model, perhaps just different constants). Some the 1960's models have 11 transitions. The main discrepancies in terms of absolute (versus relative) altitude are for the outer transitions, where the atmosphere is thin. What I don't recall or know is how much difference in drag or heat variation this causes, but as posted by boneh3ad, they compensated with extra margin in heat handling, and attitude control (such as small thrusters).
Also makes sense. So that's how youd do it for building your ship, how would you adjust your model? Say for example you flew it a couple times and in the upper transitions, in the upper atmosphere it was always 30% off. How would you adjust your model?
 
  • #20
Ultimately I would imagine that they base initial calculations on the standard atmosphere, but when it comes time for the actual reentry to occur, they rerun the numbers beforehand using current measured atmospheric conditions (wind, temperature, humidity, etc) and get a pretty good estimate from that.
 
  • #21
I found this https://en.wikipedia.org/wiki/Atmospheric_model which I thought talked about using adjustable constants. I also found this https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19870001230.pdf which talks about using a scale factor to adjust it, and I found this https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20100002996.pdf which talks about adding a density factor to adjust it on a Mars model. It seems to me like you run it with the current conditions first, and besides that it seems really similar to the aerodynamics, using constants and correction factors to tune it. Do any of you guys who know the aerodynamics/atmosphere stuff have any thoughts on that?
 
  • #22
Why are you asking us if you just linked to a report by actual engineers who seemingly actually do these sort of analyses?
 
  • #23
I mean you actually work in experimental aerodynamics so you understand it way better than me, I want to make sure I am understanding what I am reading right
 
  • #24
Sure but I don't calculate reentry trajectories. I'd have to read the paper same as you to understand what it says.
 
  • #25
Im just looking for someone to tell me it seems to them like that's what theyre saying in those papers too, or mabey, I think its actually something else. If someone who gets aerodynamics or even modeling well tells me yeah, it seems to them like that's what theyre saying too, then I am pretty sure that it is. Just some confirmation that it seems like that's what theyre saying to someone else too, or mabey it seems like theyre saying something else, so I know that I am sort of on point
 
  • #26
Regardless of what their models start with -- theoretical equations or wind tunnel results, they would be very careful for the first few flights. They would take the results of the first flights and make adjustments to the models till they get the same model answers as the actual flight data. I think the adjustments are as much an art as it is a science. The tables of aerodynamic numbers are large and have to fit together consistently. You have to understand that the people working on the aerodynamics of any vehicle have lived and breathed those numbers for years. I don't know if even they could give you a brief, methodical explanation of their process.
 
Last edited:
  • #27
The problem is you seem to be continually asking essentially the same question in different ways and about different papers. When it comes to aerodynamics problems like this, the same general process applies regardless of the model that is being used. Essentially, the fluids problem could be solved "exactly" with the Navier-Stokes equations, but that is far too difficult to be practical, so various forms of models may be used, whether it something relatively high fidelity like a turbulence model or something low fidelity like using a drag coefficient. The process from there is relatively similar across the various forms of modeling: come up with a model, test it against experiment, change the model or use a correction factor of some kind, and then retest the model.
 
  • Like
Likes FactChecker
  • #28
aero, chiming into agree with boneh3ad, who has been very patient with your questions. If you're really interested, you first might learn the basics of aerodynamics and the modeling of, which can take a lot of reading and math and possibly college courses. Luckily, that can be a fascinating subject, but you can't learn it on a forum.

Maybe you can find a local science teacher to discuss this with in person and work through some of the basics? And apologies, that all said, I don't have a suggestion on a good re-entry aerodynamics textbook.
 
  • #29
FactChecker said:
Regardless of what their models start with -- theoretical equations or wind tunnel results, they would be very careful for the first few flights. They would take the results of the first flights and make adjustments to the models till they get the same model answers as the actual flight data. I think the adjustments are as much an art as it is a science.
That makes sense. I didnt even think of that.
boneh3ad said:
Essentially, the fluids problem could be solved "exactly" with the Navier-Stokes equations, but that is far too difficult to be practical, so various forms of models may be used, whether it something relatively high fidelity like a turbulence model or something low fidelity like using a drag coefficient. The process from there is relatively similar across the various forms of modeling: come up with a model, test it against experiment, change the model or use a correction factor of some kind, and then retest the model.
This really cleared up a lot of confusion on it. Thanks for the help with this stuff. Extremely helpful.
Im going to go with this
aero5682 said:
I found this https://en.wikipedia.org/wiki/Atmospheric_model which I thought talked about using adjustable constants. I also found this https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19870001230.pdf which talks about using a scale factor to adjust it, and I found this https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20100002996.pdf which talks about adding a density factor to adjust it on a Mars model. It seems to me like you run it with the current conditions first, and besides that it seems really similar to the aerodynamics, using constants and correction factors to tune it. Do any of you guys who know the aerodynamics/atmosphere stuff have any thoughts on that?
since I am pretty much sure this is the right way to interpret it now and get the process and stuff. If anyone knows this specifically or does this, and that's off let me know, it would be helpful, though I am sure this is the right interpretation now
 
  • #30
boneh3ad said:
Essentially, the fluids problem could be solved "exactly" with the Navier-Stokes equations, but that is far too difficult to be practical, so various forms of models may be used, whether it something relatively high fidelity like a turbulence model or something low fidelity like using a drag coefficient.
I agree with this and think it is worth emphasizing. I have seen programs where there was a lot of money involved, but they still were not able to use the most accurate models. The main reason is that there are so many conditions that have to be studied. For an airplane design, the aerodynamics have to be calculated for all possible combinations of angles of attack and sideslip, air densities, roll, pitch, and yaw rates, and surface positions. That adds up to many thousands (maybe millions?) of combinations to consider. In those programs, using the most accurate model for each case is inconceivable.
 
Last edited:
  • #31
Thats really interesting, and helpful. I know from making planes in simulations it seems like its all about balance and tradeoffs. I couldn't imagine working it out for every situation. It seems like the best you can do is do it the best you can and then just try it out
 

1. What is reentry aerodynamics and why is it important?

Reentry aerodynamics is the study of the forces and behaviors that occur when an object or spacecraft reenters Earth's atmosphere. It is important because it affects the safety, stability, and trajectory of the object during reentry, which is a critical phase of space missions.

2. How do scientists determine the reentry aerodynamics characteristics of a spacecraft?

Scientists use various methods such as wind tunnel testing, computer simulations, and flight experiments to gather data and analyze the aerodynamics characteristics of a spacecraft during reentry. They also take into account factors such as the shape, size, and materials of the spacecraft.

3. What are the main challenges in studying reentry aerodynamics?

One of the main challenges in studying reentry aerodynamics is the extreme conditions that occur during reentry, such as high temperatures, pressures, and speeds. These conditions can be difficult to replicate in experiments and require advanced technologies and techniques to accurately measure and analyze.

4. How do reentry aerodynamics characteristics differ for different spacecraft?

The reentry aerodynamics characteristics can vary greatly depending on the shape, size, and materials of the spacecraft. For example, a capsule-shaped spacecraft will experience different aerodynamic forces compared to a winged spacecraft. The angle of reentry and the speed of the spacecraft also play a significant role in determining its aerodynamics characteristics.

5. How do the reentry aerodynamics characteristics affect the design of a spacecraft?

The reentry aerodynamics characteristics play a crucial role in the design of a spacecraft, as it determines the structural integrity, stability, and control of the spacecraft during reentry. Engineers must carefully consider the aerodynamics forces and behaviors in order to design a spacecraft that can safely and successfully reenter the Earth's atmosphere.

Similar threads

Replies
8
Views
2K
Replies
1
Views
1K
  • Mechanical Engineering
Replies
2
Views
1K
  • Mechanical Engineering
Replies
1
Views
1K
  • Mechanical Engineering
Replies
1
Views
3K
Replies
1
Views
3K
  • Mechanical Engineering
Replies
1
Views
131
  • Electrical Engineering
Replies
2
Views
301
  • General Engineering
Replies
1
Views
3K
  • Classical Physics
Replies
30
Views
3K
Back
Top