Observations of drag at the back of craft

[Foxster]

See, I thought drag was going to be all about making sure that the front of your craft was pointy. Seems the back is just as important.

At 1km on this craft the fuel tank has a drag of 13.03...

Stick a nose cone on the bottom and the tank drag is now 1.54 but the cone has a drag of 7.9, total drag of 9.44 and carrying an extra 0.2t.

Swapping out the tail streamlining for an adapter and small nose cone, the tank drag is about the same (1.56) and the total drag of the three parts is 6.6, with 0.13 extra mass.

So, the shape of the back end does have a very significant effect on drag, especially of the parts above it. I haven't figured out what is optimal yet though.

Launching the first and last craft straight up until out of fuel revealed the lowered drag has made a difference despite the extra mass:

No back end streamlining: 36,031km @ 1095.5m/s

With back end streamlining: 37,162km @ 1117.3m/s

Anyone else optimised a rear end? (*snarf snarf*)

Edited May 14, 2015 by Foxster

[cybersol]

Have your tried the tail cone, as they work surprising well on the front if they survive the heat?

[gmiddlemass]

Very interesting discovery...

[Foxster]

Have your tried the tail cone, as they work surprising well on the front if they survive the heat?

They are quite heavy. Just tried this:

Only made it to 23km before it was out of juice. So, not so good.

[Roflcopterkklol]

Personally i found making rockets pointy stops mattering at a certain TWR, you can make some silly things fly in the new aero which is kinda nice if im honest.

Admitted that rocket does have a bug where the launch clamps smash into it at 19km and it explodes, but still it flies until then.

But on a less silly note the rocket on top of that rocket would not fly IRL.

Edited May 14, 2015 by Roflcopterkklol

[Wjolcz]

Nice discovery. I always keep the back pointy like real high-speed aircraft do, but it's good to know it actually helps like it should.

[Cuky]

See, I thought drag was going to be all about making sure that the front of your craft was pointy. Seems the back is just as important.

Well, when an object (in this case a rocket) passes through the air it pushes air around itself leaving a zone of low pressure behind. Now, airstream that is at higher pressure is coming of from the edge and naturally mixes with that zone of lower pressure and that creates vortices which in turn drags on the object and slows it down. When you have back end tapering like that of a nose cone mixing of those low and high pressure streams is much more gradual, vortices are less powerful and are causing less drag on your vehicle.

[Superfluous J]

Well, when an object (in this case a rocket) passes through the air it pushes air around itself leaving a zone of low pressure behind. Now, airstream that is at higher pressure is coming of from the edge and naturally mixes with that zone of lower pressure and that creates vortices which in turn drags on the object and slows it down. When you have back end tapering like that of a nose cone mixing of those low and high pressure streams is much more gradual, vortices are less powerful and are causing less drag on your vehicle.

The surprise here (for me at least) was that the game actually models (or at least approximates) this. I, too, assumed all that mattered was the front of your rocket.

[Sky_walker]

You tried swapping your radial engines with a single rear engine with the same thrust and testing how that affects drag? (Gases from engine should fill the vacuum behind the rocket reducing the drag - or at least: they would in real life)

Is drag of your service bay affected in a same way as the fuel tank? (Makes me curious if it's just a last part that has an increased drag, or all of them)

You tried adding two smaller fuel tanks instead of one larger and seeing how this affects drag - outcome total drag for the rocket should be identical, but is it?

Edited May 14, 2015 by Sky_walker

[Foxster]

Well, when an object (in this case a rocket) passes through the air it pushes air around itself leaving a zone of low pressure behind. Now, airstream that is at higher pressure is coming of from the edge and naturally mixes with that zone of lower pressure and that creates vortices which in turn drags on the object and slows it down. When you have back end tapering like that of a nose cone mixing of those low and high pressure streams is much more gradual, vortices are less powerful and are causing less drag on your vehicle.

Well, yes, I did realise this but, like above, I'm surprised it is modelled.

- - - Updated - - -

You tried swapping your radial engines with a single rear engine with the same thrust and testing how that affects drag? (Gases from engine should fill the vacuum behind the rocket reducing the drag - or at least: they would in real life)

Is drag of your service bay affected in a same way as the fuel tank? (Makes me curious if it's just a last part that has an increased drag, or all of them)

You tried adding two smaller fuel tanks instead of one larger and seeing how this affects drag - outcome total drag for the rocket should be identical, but is it?

The combined drag of two smaller tanks seems to be about that of one equivalent larger one.

The high drag gets pushed to the back no matter how many parts I added.

I'll have a play with what a single engine does but they aren't so easy to play with as the radials. I'll bet though that an engine that's the same size as the tank it is fitting to will result in lower overall drag than a mismatch.

Update...Checked and engines do act reduce the drag of the parts above but then so does any size-matched part. What makes the difference to the overall drag is how draggy that last part is and the bigger engines are very draggy.

Edited May 14, 2015 by Foxster

[Michaelo90]

Well, yes, I did realise this but, like above, I'm surprised it is modelled.

- - - Updated - - -

The combined drag of two smaller tanks seems to be about that of one equivalent larger one.

The high drag gets pushed to the back no matter how many parts I added.

I'll have a play with what a single engine does but they aren't so easy to play with as the radials. I'll bet though that an engine that's the same size as the tank it is fitting to will result in lower overall drag than a mismatch.

Update...Checked and engines do act reduce the drag of the parts above but then so does any size-matched part. What makes the difference to the overall drag is how draggy that last part is and the bigger engines are very draggy.

I tested this with a 2,5 m rocket with both a skipper and a mod 1,25 m engine, with the skipper thrust limited to the same sea level thrust. The 1,25 m version reached 26 km while the skipper version reached 31 km, despite being 1t heavier, so having the right engine size seems to matter a LOT.

[magnemoe]

You tried swapping your radial engines with a single rear engine with the same thrust and testing how that affects drag? (Gases from engine should fill the vacuum behind the rocket reducing the drag - or at least: they would in real life)

Is drag of your service bay affected in a same way as the fuel tank? (Makes me curious if it's just a last part that has an increased drag, or all of them)

You tried adding two smaller fuel tanks instead of one larger and seeing how this affects drag - outcome total drag for the rocket should be identical, but is it?

This, engine does not even have to produce useful trust. http://www.globalsecurity.org/military/systems/munitions/base-bleed.htm

Not sure if this work in KSP.

[monophonic]

Nice find. I do wonder though whether the destabilizing effect of boattails is modeled as well. I'm on the move so can't test myself right away.

[Temeter]

That's interesting. I usually just put some detachable caps below rcokets for the visuals, didn't expect them to be actually useful.

This, engine does not even have to produce useful trust. http://www.globalsecurity.org/military/systems/munitions/base-bleed.htm

Not sure if this work in KSP.

Tried it, sadly not. Although a rocket nozzle probably wouldn't completely mitigate the low pressure area.

Edited May 14, 2015 by Temeter

[Sky_walker]

I tested this with a 2,5 m rocket with both a skipper and a mod 1,25 m engine, with the skipper thrust limited to the same sea level thrust. The 1,25 m version reached 26 km while the skipper version reached 31 km, despite being 1t heavier, so having the right engine size seems to matter a LOT.

Good Nice work testing it.

Seems like it works better than I expected.

[Captain Sierra]

How punishing are clustered engines now?

[Temeter]

Btw, an observation: Rockets with cones on their underside are alot more stable and less likely to tumble. Drag difference can be huge btw, especially when it comes to high t/w ratios and supersonic below 10k.

Another result, putting a nosecone below the rocket and using radial engines is more stable than using an inline engine.

Edited May 14, 2015 by Temeter

[CocoDaPuf]

Hey Foxter, great post! I'd also love to see how what drag is like with the new fairings. Could you try this again with the same configuration as you have in the original post, but using a shaped fairing on the tail piece? Maybe once with a sharper point and one rounder?

[Foxster]

Hey Foxter, great post! I'd also love to see how what drag is like with the new fairings. Could you try this again with the same configuration as you have in the original post, but using a shaped fairing on the tail piece? Maybe once with a sharper point and one rounder?

Already tried it. Just inverted it and built down.

It was poor. The drag of it is relatively high and the extra mass negates any other usefulness. It made my little test rocket perform worse than putting nothing on the back end.

[CocoDaPuf]

Wow, pretty lame. But good to know!

- - - Updated - - -

Tried it, sadly not. Although a rocket nozzle probably wouldn't completely mitigate the low pressure area.

Well, wouldn't an active rocket (more than) completely mitigate that low pressure area? I mean, rockets generate thrust by expelling gas at high pressure, that's the whole idea right?

[Temeter]

Well, wouldn't an active rocket (more than) completely mitigate that low pressure area? I mean, rockets generate thrust by expelling gas at high pressure, that's the whole idea right?

The exhaust would only cover a small cone behind the tip of the nozzle. It would need to cover the whole backside of the rocket to fully migitate the low pressure area.

That's just logical assumptions tho, might be complete nonsense.

Edited May 14, 2015 by Temeter

[Enorats]

Discoveries like this should grant science points.

- - - Updated - - -

Holy difference too, WOW! Sent a rocket similar to those the OP made straight up and it hit an apoapsis of 95k before running out of fuel. Maxed out at around 950 m/s too. Then I sent another with the two nose cones on the bottom like the OP did, and it maxed out at 305k and around 1300 m/s. Crazy.

Can't really test how a rocket engine would affect it sadly, as there are none that come close to matching the radials in TWR/Isp and the drag created by the radials is also an issue. The skipper seemed the closest match, but that required throwing fins on in place of the radials and limiting the thrust to 73% to match the TWR of the radials. Still not really a good match though, as there are too many variables. It got about 3 times higher than even the aerodynamic radial version.

- - - Updated - - -

Interestingly, using a fairing in place of the two nose cones on the radial got me to 315k.. even though the fairing is about twice the mass of the two nose cones.

- - - Updated - - -

Odd.. jettisoning the fairing around 30k got the fairing version to 350k apo. I thought jettisoning them made no difference in vessel mass...

- - - Updated - - -

Hmm.. multiple trials of the above show there to be a large degree of error in the results (upwards of 10%), even when the rockets are being fired straight up with no control input whatsoever. No idea why that would be.

[Temeter]

Odd.. jettisoning the fairing around 30k got the fairing version to 350k apo. I thought jettisoning them made no difference in vessel mass...

Fixed in 101.

[NathanKell]

It's also worth noting that rear drag tapers off with mach: while subsonic, the dragginess of the rear of your vessel matters a lot, whereas at high mach it doesn't matter very much at all.

By definition, if the rear of your craft is less draggy, your craft is less stable. So the destabilizing effects of boattails are modeled. Recall that stability lies in having the CoM above the aerodynamic center; if you decrease the drag of the tail of the vessel, that moves the aerodynamic center up (because, well, the AC is the center of all aerodynamic forces, and you just decreased the forces at the bottom).

[mikegarrison]

Well, when an object (in this case a rocket) passes through the air it pushes air around itself leaving a zone of low pressure behind. Now, airstream that is at higher pressure is coming of from the edge and naturally mixes with that zone of lower pressure and that creates vortices which in turn drags on the object and slows it down. When you have back end tapering like that of a nose cone mixing of those low and high pressure streams is much more gradual, vortices are less powerful and are causing less drag on your vehicle.

That's mixing two different ideas. It doesn't require any large-scale vortices for this effect to occur.

What you are trying to describe is called "form drag" or "pressure drag". When you push a solid object through the air, the air in front gets squished, giving it a high pressure. In a completely inviscid flow, this pressure is recovered at the back, and the net effect is no drag. But real air is not inviscid. In real air, the flow does not completely fill in at the back. Instead, a boundary layer forms around the object, and this boundary layer prevents the freestream flow from completely closing at the back. So you have low pressure at the back and high pressure at the front. The difference between the two is a net force from the pressure: drag.

The more that you leave a blunt surface at the back, the more that you get a big area of low pressure pulling back on you. Or another way to put it, the bigger of a hole that you punch through the air. A shape like an airfoil, however, gives the air a chance to fill in behind your object without leaving a huge hole of low pressure. That's why an airfoil shape has less drag than other shapes. It's also why airfoils have a sharp and pointing trailing edge.

Edited May 14, 2015 by mikegarrison