Rocksim: why do Conical nose cones sim higher than Von Karman nose cones

[ghostfather]

I'm simming a 75mm minimum diameter design around a CTI M2245 Imax motor. It should reach speeds between Mach 2.6 and 2.9.
When I only change the shape of the nose cone between Conical and Von Karman, it makes about a 30% difference.

Conical: Mach 2.86, 15138 m high
Von Karman: Mach 2.64, 10680 m high

Sure, the wetted surface is less on the conical nose cone compared to the Von Karman, but I would assume that Rocksim also factors in other things like shock waves produced by the conical nose cone.


I often refer to this diagram when choosing a nose cone shape, and the Von Karman shape seems to be the best all-round shape, and better than a Ogive or Conical.

Can anyone tell me what's going on in Rocksim, why the huge difference? I've noticed that Rocksim also favors Conical nose cones in lower speed flights as well, under Mach 1.5. It seems improbable that Conical shapes would outperform a Von Karman shape.
Of course, I could try it in OpenRocket or RASAero.

 

[jmasterj]

You need to use RASAero for accuracy in high-mach flights.

 

[cls]

Good of you to notice. It's an old problem, built in to the DATCOM model. So yes, conical nose cones are not as good as they sim out.

Proof: 2 flights of my Mongoose 98, on CTI N1100. One with conical, one with Von Karman nose cone. VK flight went over 10% higher and faste: Accel and thrust curve of both motors matched within a couple percent. All up launch weights were almost equal.

 

[SolarYellow]

I hypothesize that on real sounding rockets, the reason they use conical is the fabrication resources to spin an ogive or VK cone simply aren't available. Spinning large items like that from aluminum is pretty specialized. I've read that there was this one guy with the capability (fabrication skill/knowledge/equipment) to make the ogive cones for the Aerobee, but who knows if that's really the case.

 

[Adrian A]

ell me what's going on in Rocksim, why the huge difference? I've noticed that Rocksim also favors Conical nose cones in lower speed flights as well, under Mach 1.5. It seems improbable that Conical shapes would outperform a Von Karman shape.
Of course, I could try it in OpenRocket or RASAero.

This is a known issue in RockSim that I pointed out to them in 2011. When they never fixed it I stopped buying it. RASAero is really the gold standard for sim accuracy for high performance flights, and even OpenRocket is closer than RockSim (edit: for supersonic flights).

 

[SolarYellow]

OpenRocket has a problem with elliptical nose cones due to the incompleteness of data in the literature. It massively overestimates the drag of an elliptical nose cone in any speed range in which elliptical would be likely to be the best performer in reality. But at least I don't have to pay for something I know gets me the wrong answers in a wide range of cases I'm interested in.

 

[watheyak]

I had heard the noseconedrag model is based on total surface area. Since the conical has the least surface area, Rocksim and OpenRocket show it as the least draggy.

I'll also +1 the use of RASAero for anything above the speed of sound.

 

[neil_w]

I had heard the noseconedrag model is based on total surface area. Since the conical has the least surface area, Rocksim and OpenRocket show it as the least draggy.

Not true, at least in the tests I just did with the current version of OR. I just played around and it showed conical as the draggiest of all the shapes. Higher pressure drag. Try it for yourself, using Tools -> Component Analysis, "Drag Characteristics" tab.

There is still some work to do to improve the nose cone drag model, though, no doubt.

 

[user 30299]

OpenRocket has a problem with elliptical nose cones due to the incompleteness of data in the literature. It massively overestimates the drag of an elliptical nose cone in any speed range in which elliptical would be likely to be the best performer in reality.

How did you determine it massively overestimates the drag?

 

[rocket_troy]

I had heard the noseconedrag model is based on total surface area. Since the conical has the least surface area, Rocksim and OpenRocket show it as the least draggy.

That's wrong. It's not based on total surface area - obviously, otherwise we'd be using flat plates as NCs.

TP

 

[watheyak]

That's wrong. It's not based on total surface area - obviously, otherwise we'd be using flat plates as NCs.

TP

I'm saying in the sim that's the case, not reality. In Rocksim a flat plate actually sims higher than anything else because of this. Or at least it used to. It was one of the first things I tested when 10 came out.

I'm glad that OpenRocket has updated the model, as @neil_w pointed out. Time to upgrade!

 

[rocket_troy]

I'm saying in the sim that's the case, not reality. In Rocksim a flat plate actually sims higher than anything else because of this. Or at least it used to. It was one of the first things I tested when 10 came out.

I'm glad that OpenRocket has updated the model, as @neil_w pointed out. Time to upgrade!

The only reason I mentioned it, is that I have heard some people on other forum/lists make quite adamant assertions that surface area was the fundamental driver and that's been a bit of a bugbear of mine since.

TP

 

[SolarYellow]

How did you determine it massively overestimates the drag?

Hashed it out over several discussions here over time with the current developers of OR. Basically, the pressure drag of most nose cones is very low (but with elliptical demonstrated to be the best) until near the transonic range, where it begins rising. Most of them have a bump through the transonic, then decrease again at supersonic. VonKarman's "bump" is the least-high in that range. Elliptical tends to be the highest. The problem is that the graph in the literature that actually provides CDs for the various shapes, and which OR bases its form drag calculations on, doesn't show anything for elliptical cones below Mach 1.2, where the CD is ~0.11. So OR just linearly interpolates between that point and zero pressure drag at Mach 0.39 for ellipticals, whereas all the other shapes have data and it is able to use the curves that are very low for most of the speed range from zero up to near transonic. I've found other sources for CD of an elliptical nose cone, but none that included CD vs. speed.

https://www.rocketryforum.com/threads/elliptical-nose-cones-in-openrocket.174284/#post-2312268

 

[user 30299]

Not true, at least in the tests I just did with the current version of OR. I just played around and it showed conical as the draggiest of all the shapes. Higher pressure drag. Try it for yourself, using Tools -> Component Analysis, "Drag Characteristics" tab.

There is still some work to do to improve the nose cone drag model, though, no doubt.

Does your nose cone drag calcs take into account the Fineness Ratio? (Nose Length to Nose Base Dia.; L/D)

In the early research the nose cone's fineness ratio was a major determinant (influence) on nose drag.
This was in speeds below the Transonic range. The reports show 3.0 was a favored ratio for testing.
The ratio was tied back to Skin Friction Drag.

The chart in the #1 Posting is based on nose cones that had a Fineness Ratio of 3.0.

 

[neil_w]

Does your nose cone drag calcs take into account the Fineness Ratio? (Nose Length to Nose Base Dia.; L/D)

I'll defer this to @JoePfeiffer.

 

[JoePfeiffer]

Does your nose cone drag calcs take into account the Fineness Ratio? (Nose Length to Nose Base Dia.; L/D)

Yes. From the comments in .../core/src/net/sf/openrocket/aerodynamics/barrowman/SymmetricComonentCalc.java

* First, the transonic/supersonic region is computed. For conical and ogive shapes
* this is calculated directly. For other shapes, the values for fineness-ratio 3
* transitions are taken from the experimental values stored above (for parameterized
* shapes the values are interpolated between the parameter values). These are then
* extrapolated to the current fineness ratio.

 

[user 30299]

Yes. From the comments in .../core/src/net/sf/openrocket/aerodynamics/barrowman/SymmetricComonentCalc.java

* First, the transonic/supersonic region is computed. For conical and ogive shapes
* this is calculated directly. For other shapes, the values for fineness-ratio 3
* transitions are taken from the experimental values stored above (for parameterized
* shapes the values are interpolated between the parameter values). These are then
* extrapolated to the current fineness ratio.

Thanks for the info!

 

[pugachu]

The chart in the #1 Posting is based on nose cones that had a Fineness Ratio of 3.0.

Is 3.0 fineness ratio for a Von Karmen the best that you can get for a VK profile, or does it optimize at a different value than an ogive or conical nose?

I'm asking, because I need to make a new mold for the next iteration of my 54mm min diameter rocket. my current one is closer to a 4:1, but the shoulder is made for thin wall CF tubing.

 

[user 30299]

Is 3.0 fineness ratio for a Von Karmen the best that you can get for a VK profile, or does it optimize at a different value than an ogive or conical nose?

I'm asking, because I need to make a new mold for the next iteration of my 54mm min diameter rocket. my current one is closer to a 4:1, but the shoulder is made for thin wall CF tubing.

I do not have an answer for your VK fineness ratio (FR) question.

There is a smattering of reports, spread over the years, that look at the FR for
various cone shapes - but there is no definitive chart or graph that you can just
look at and there's your answer. At least I have not found one yet.

It's like the information is there, but you would have to sit down and pull it all
together to make a reasonable "postulation" of the best FR for a nose cone.

I have seen in a couple of reports the statement that once you reach a FR of 6,
the cone's impact on drag is negligible. It also has been noted that as the ratio
increased the longitudinal stability of the nose cone & body decreased.

Attached are two reports that may help you out, or at least give you a little
more background on FR.

 

[pugachu]

I do not have an answer for your VK fineness ratio (FR) question.

There is a smattering of reports, spread over the years, that look at the FR for
various cone shapes - but there is no definitive chart or graph that you can just
look at and there's your answer. At least I have not found one yet.

Thanks for the info. I always like learning new stuff. I didn't think there was an easy answer, but i was hoping there would be.

 

[Adrian A]

Thanks for the info. I always like learning new stuff. I didn't think there was an easy answer, but i was hoping there would be.

I have been using VK cones with fineness ratio of 7 in 38mm record attempt rockets and getting lower Cd than predicted in supersonic flights. I think higher fineness ratio works better on faster flights.

Reading the first paper is interesting. If you pay close attention to the scales you can see that the increase in skin friction drag for higher fineness ratios is tiny compared to the wave drag. Unfortunately, the plot scale for wave drag is very compressed so it’s hard to quantify from those graphs.

 

[user 30299]

I have been using VK cones with fineness ratio of 7 in 38mm record attempt rockets and getting lower Cd than predicted in supersonic flights. I think higher fineness ratio works better on faster flights.

Reading the first paper is interesting. If you pay close attention to the scales you can see that the increase in skin friction drag for higher fineness ratios is tiny compared to the wave drag. Unfortunately, the plot scale for wave drag is very compressed so it’s hard to quantify from those graphs.

In another thread, Joe Pfeiffer has pointed to the following report, NASA-TR-R100, as a relevant source.
It's attached in this post. It's chock full of drag information in good summaries.

 

[user 30299]

Thanks for the info. I always like learning new stuff. I didn't think there was an easy answer, but i was hoping there would be.

If you want to dig a little deeper into the drag and fineness ration for VK cones, here is a link for
downloading NACA Tech Report 1386; https://ntrs.nasa.gov/citations/19930091022

The report was too big to download on TRF.

The report will also introduce you to how the nose bluntness can affect the cone's drag.
I found that pretty interesting.

There are also reports out there that look at the effects of the "afterbody" elements on
the nose drag. Afterbody meaning the length of the body tube (fuselage), fins, etc.

Many of the reports are from the 1950's, are helpful in understanding drag, but I don't
know well they relate to our era of CFD. That's beyond my skill set. lol

 

[Adrian A]

In another thread, Joe Pfeiffer has pointed to the following report, NASA-TR-R100, as a relevant source.
It's attached in this post. It's chock full of drag information in good summaries.

Wow, there's a lot in there.

 

[SolarYellow]

CFD analysis of VK and LD-Haack nose cones of fineness ratios 3, 4, and 5 through the speed range most people involved in this conversation probably care about the most. Direct link to PDF download of paper. Chrome might require you to copy the link and paste it into your browser bar.

https://www.uah.edu/images/administrative/Honors/Papers/Chad_OBrien.pdf
One detail to note is that the viscous drag is significant, especially for the larger fineness ratio. It's generally greater than pressure drag until about Mach 0.9 or so. Even at lower speeds where a smaller fineness ratio NC might look better, the drag for a smaller fineness ratio NC plus a straight section of BT adding up to the same length as a greater fineness ratio NC will be greater than for the NC of greater fineness ratio. So if you can use the length of the greater fineness ratio nose cone to stuff some recovery gear or electronics up there and shorten the airframe corresondingly, you'll be ahead. At least one of the guys doing an H13 record attempt basically made his whole forebody a very long VK nose cone, from what I could tell in his pics. I think that's the way to go, as much as you can pull it off.

 

[user 30299]

CFD analysis of VK and LD-Haack nose cones of fineness ratios 3, 4, and 5 through the speed range most people involved in this conversation probably care about the most. Direct link to PDF download of paper. Chrome might require you to copy the link and paste it into your browser bar.

https://www.uah.edu/images/administrative/Honors/Papers/Chad_OBrien.pdf

Thanks for the link. It worked fine. Looks like a good read.

 

[user 30299]

CFD analysis of VK and LD-Haack nose cones of fineness ratios 3, 4, and 5 through the speed range most people involved in this conversation probably care about the most. Direct link to PDF download of paper. Chrome might require you to copy the link and paste it into your browser bar.

https://www.uah.edu/images/administrative/Honors/Papers/Chad_OBrien.pdf

Read the paper. Pretty good mini thesis. O'Brien takes his CFD findings and relates them back
to one of the early NACA reports on drag.