Notchbell and the Boot compensators

guest blogger john hare

Compensating nozzles would be useful if they can be done affordably. Affordable is not only the cost to fabricate a nozzle, but also the risk of pursuing things that don’t work, and the time and money to develop it to operational status. Years behind the competition due to excessive chasing chimeras of perfect performance can be more expensive than just taking a performance hit. An affordable compensating nozzle should be something that can be modified from existing equipment and brought to use in minimum time with minimal cost.

One simple and cheap method that shouldn’t require much in resources to build or test is what I call a notchbell nozzle. It is a large expansion ratio bell nozzle with a notch in it up to the plane of the lowest expansion ratio desired by the flight profile. It works in the same manner as a linear spike nozzle except there are no other combustion chambers. The exhaust of the single throat expands against the normal bell on all sides except the atmospheric wall on the notch side. The air controls the plume in the same manner as the spike nozzle that has metal reaching all the way down to the maximum expansion ratio with no aero component to the spike.

The engine will have asymmetrical thrust and need more gimbal authority than normal.

 

The two sketches here represent two different concepts. The one on the left is the notchbell. The one on the right is a speculative bolt on adapter to get compensation from an unmodified bell nozzle.

When trying to figure out what went wrong with some notchbell units I made, I remembered some things from an intake aerodynamics book. Subsonic intakes on jet aircraft have smooth rounded leading edges of fairly precise design to maximize air pressure to the engine with the minimum reasonable drag. They do not have sharp edges like supersonic intakes. As far as I can tell, it might be possible to add a tiny rounded intake to an over expanded bell nozzle to force it to separate in a planned way to compensate for pc/pa variations. The boot would be on the cool side of the nozzle until it reached the nozzle design expansion altitude. It could be made of light expendable material that burned off after doing its’ job.

Since the boot would only be used on about 25% of the perimeter, it would create asymmetrical thrust by forcing flow separation on just part of the normal nozzle. If the boot is found to work, then two of them could be placed on opposite sides of the nozzle to make the thrust symmetrical. It is possible that some TVC can be done with four boots sliding in and out of position. Once the engine is in vacuum, normal means of TVC must be used as there will be no separation to control.

I made a tiny notchbell unit that hooked into shop air a few years back. The first one worked exactly like it was supposed to with the air flow focused on the wall right down the center line opposite the notch. Then I promptly destroyed any credibility the thing might have had by whipping out a dozen of them and shipping them off untested to friends in the business. I only realized the problem when I did another batch of two dozen and only three of them worked properly the first time. I had to spend a good bit of time tweaking them to get up to nineteen units that worked reasonably well. I am hoping some of the people that got the stupid things will take another look.

The boot is highly speculative. The only saving grace it has is that it would be incredibly fast and cheap to confirm or bust. No major hardware would have to be damaged to test the concept. It is almost to the point of shaping a short length of round stock to fit 25% of the perimeter of your large expansion nozzle, and holding it on with a hose clamp during a characteristics burn. Maybe a little caulk to keep a smooth joint pipe to nozzle. It is possible that a couple hours construction and 4 seconds of burn could tell the tale.

This sketch may help clarify what I suggested in comments as a visual aid in understanding the boot idea. A pipe connects the ambient air to the low pressure center in an over expanded bell. Air should flow from the higher pressure ambient to the low pressure center cutting down losses. Putting a pipe in the exhaust stream like this would melt it. The boot is an attempt to get the same job done without interfering with the exhaust stream.

The following two tabs change content below.
johnhare

johnhare

I do construction for a living and aerospace as an occasional hobby. I am an inventor and a bit of an entrepreneur. I've been self employed since the 1980s and working in concrete since the 1970s. When I grow up, I want to work with rockets and spacecraft. I did a stupid rocket trick a few decades back and decided not to try another hot fire without adult supervision. Haven't located much of that as we are all big kids when working with our passions.
johnhare

Latest posts by johnhare (see all)

This entry was posted in Uncategorized. Bookmark the permalink.

25 Responses to Notchbell and the Boot compensators

  1. Tim says:

    Its entirely possible that I don’t understand the notchbell, but could you use two notches at 180 degrees to get symmetrical thrust? Or three at 120 degrees?

  2. Habitat Hermit says:

    Same as Tim and how much would it cost for example MSS to detail and test this?

    And great band name by the way ^_^

  3. john hare says:

    Tim,

    If you mean multiple chambers, then the saddlespike has the two notches with the connecting ‘saddle’ to save length and mass. Triangle three and phillips four and so on until an honest annular aerospike is reached are equally possible.

    If you mean notching the bell of one engine multiple times, then I think it is possible, just more difficult to figure out. I worry about structural mass and strength with multiple thin cantilever nozzle sections. This will probably take some CFD work to figure out.

    The aerospike nozzles in general are counter intuitive and involve development risk technically and financially. If they were a well understood method with substantial known benefits, Falcon 9 would almost certainly use them with those nine seperate engines. My intent is to find a financially safe method of reaping the technical benefits of the aerospike method.

  4. john hare says:

    Habitat,

    I think about $10k to do an initial reality check of the notchbell assuming that Masten was willing to bolt a test article to one of their existing engines. The boot could be a tenth of that if they have an over expanded nozzle in stock.

    My original intent from years ago was to achieve the financial ability to pay experienced people to test different concepts to find the ones that would give me serious competitive advantage. Ma nature and the test stand could judge my sanity or lack of. Now I am hoping someone will check them out in simulation or hardware on their own dime and at least mention it to the community.

  5. Eric Collins says:

    If you send me a geometry description of the nozzle and the flow conditions you’d like to run it under, and I could try to put it through a few CFD runs. I’m kinda busy at the moment trying to finish up a dissertation, but perhaps sometime in the next couple of months I could take a look at it.

  6. John,
    I’m not sure I get your boot compensator idea. But for the notchbell idea (I was one of the people you gave a copy of that fiberglass proof of concept nozzle to), I could see at least one potential problem. I think the concept would actually work at low altitude, but as the altitude increases, or chamber pressure is throttled up, eventually you’ll get flame shooting out the notch. While downstream of the notch might act like an aerospike/E-D nozzle, eventually the ambient pressure at the notch point will be lower than the exit pressure, and at that point, you’ll probably start getting substantial spillage out the notch. I would think. I could be wrong though, I’m no CFD expert (I used it once in school, but that doesn’t really make one an expert).

    On the plus side, this idea isn’t that far out there. Actually, it looks very similar to a concept that I believe was tested at some point. It used closeable ports around the nozzle that would be opened at low altitude, and would close as the altitude increased. It was described in that AIAA paper that covered different altitude compensating approaches.

    But your guess at how much it would cost to test it is probably in the right ballpark for what we’d probably charge to test out a simple concept like that. The 750 engine we’re about to start testing was designed from the start to allow us to bolt-on various nozzle attachments. [I’ve got a few ideas I am hoping I can get time to test as well.]

    ~Jon

  7. john hare says:

    Eric,

    I will try to describe and draw the thing as clearly as I can. communication is not my strong suite.

  8. john hare says:

    Jon,

    “”John,
    I’m not sure I get your boot compensator idea. But for the notchbell idea (I was one of the people you gave a copy of that fiberglass proof of concept nozzle to), I could see at least one potential problem. I think the concept would actually work at low altitude, but as the altitude increases, or chamber pressure is throttled up, eventually you’ll get flame shooting out the notch. While downstream of the notch might act like an aerospike/E-D nozzle, eventually the ambient pressure at the notch point will be lower than the exit pressure, and at that point, you’ll probably start getting substantial spillage out the notch. I would think. I could be wrong though, I’m no CFD expert (I used it once in school, but that doesn’t really make one an expert).”””

    I don’t have the quoting thing down on blogs, hope people will continue to bear with me.

    The boot is an add on to smooth the airflow into a nozzle that is seriously overexpanded. When the exhaust really wants to separate, a fairly small assist could let the 14.7 flow into the 7psi region at a predicted place, I think. When the overexpansion is to the point that separation is going to take place anyway, the boot simply directs it to a particular location. One visual aid is to picture a pipe bent into a U like a plumbing trap with one end inside the overexpanded region of the nozzle and the other in clean air. The air will flow to the lower pressure region inside the nozzle.

    The potential problem with the notch is what worries me also. I have a real problem visualizing how a linear aerospike works too. I am more or less suggesting that if the linear aerospike works, then this will. No linear aerospike has ever flown that I am aware of. If I have the concept right, it could be altitude tested on a high power amateur rocket at one of the events with a high altitude waiver. A serious amateur could really make a contribution here if it works. An alternate method of testing is to carry the nozzle to various altitudes in an aircraft fitted with a test stand. If it worked as planned at 0-40,000 feet, then it would likely work at 100,000 feet and up.

    Another alternate method is to bolt a short notch unit to a very high pressure engine. If the engine optimised at er=20 with the notch at er=3, then it would be known what happens when an engine with 10% of that chamber pressure is at ~60,000 feet.

    Henry mentioned that the closable flap had been tried when I suggested something like that on usenet. Said it sorta worked.

    “””But your guess at how much it would cost to test it is probably in the right ballpark for what we’d probably charge to test out a simple concept like that. The 750 engine we’re about to start testing was designed from the start to allow us to bolt-on various nozzle attachments. [I’ve got a few ideas I am hoping I can get time to test as well.]””””

    I was trying for the Carmack funding method with my construction business. If it had worked out, my intention had been to run a dozen concepts at a time on a known thrust chamber by competent people. According to one source, the group discount could be substantial. It is often cheaper to write a check than to relearn why you don’t do the same stupid rocket stunt that hasn’t worked before. I am aware of your creative abilities, and why giving them away is bad business.

  9. Swatch says:

    I like the idea of the notchbell, and I like the suggestions somebody made about 3 notches at 120 degrees. Ignoring the structural consideration of a weaker bell, if the single notch works by focusing flow on the other wall then the 3 notches may offer a better balanced performance. It would be interesting to do CFD on this concept. Perhaps I shall.

  10. john hare says:

    Swatch,
    If you, Eric or somebody else develops the data for a successful compensating nozzle from these basic ideas, I have posted it on an open forum and give up any rights to the concept. If you get a chance to make money and help the industry, go for it.

    “Concepts are easy, implementation is hard.” Henry Vanderbilt

  11. Habitat Hermit says:

    “I have a real problem visualizing how a linear aerospike works too”

    The key in my opinion is to realize the importance of the difference in angles between the major axis of the vehicle and the axis of the rocket exhaust as it enters the nozzle.

    If that’s hard then imagine an ordinary rocket engine and nozzle on a horizontal test stand and then put a deflection screen behind it like for example MSS does. Then change this so that the deflection screen is fixed to and suspended by the end of the nozzle.

    Likewise in linear aerospikes the rocket exhaust hits a specialized curved deflection screen at an angle. This puts fixed pressure upon and changes the direction of the exhaust. On the other side of the exhaust is atmospheric pressure.

    On it’s own this would spin the rocket and deflection screen hence the mirroring easily seen in linear aerospikes. Annular aerospikes also mirror but they do so continuously around the major axis of the engine and nozzle combination.

    Aerospikes in general are often described as half a nozzle but this is somewhat misleading in my opinion because what you take away from one side is added to the other and angled so as to derive the most area for thrust to act upon (if one really just chopped half away one would lose half the thrust).

    However next is the realization that plasma/gas will be trapped and locked in the area of the (mirrored) spike if one removes the spike itself. It doesn’t do as well of a job in transferring thrust to the vehicle but it can save a lot of weight.

    Did that help? It is why the notch design will “bleed” while linear and annular aerospikes don’t. Not that I think such bleed will be a deal killer for all applications nor that there can’t be clever solutions to it.

    About quoting at least the way I do it (and did the above quote from you) is to use html tags. Here’s an example, first what I write:
    <i>”example”</i>
    and next how the above would be displayed:
    “example”

    “i” is the instruction to the browser to display the following in italics and it will normally continue to do so until it encounters the closing tag (think of the “/” in the last tag as a “close” command). To continue with the above example it could be read as “(start italics) example (stop italics)”.

    There’s no preview function on comments here (always helpful for checking on html use) but most places at least “i” for italics and “b” for bold works (I’ll just check).

    I might double comment because I think I started writing this hours ago (no it wasn’t written all in one go ^_^) with something else in mind but I’ve forgotten what it was…

  12. Habitat Hermit says:

    Ok “…they do so continuously around the major axis of the engine and nozzle combination.” is poorly phrased, easier to just say it’s rotated/mirrored around the axis of the direction one wants to move in.

  13. Habitat Hermit says:

    And just to avoid any misunderstandings if I had the funds I would be willing to get data on the notchbell, it’s interesting enough to merit it and get some solid data on its characteristics.

    I think mastery of controlling the flow separation is a bigger challenge for both ideas since both nozzles could easily experience big variations in stresses of the nozzle material.

  14. john hare says:

    Habitat Hermit,

    Actually quite good explanation of the linear aerospike. The notchbell units worked best with about 5 degree of bias away from the notch more or less as you describe. They were fiberglass to an air chuck. About an ounce each with a 1/16″ throat.

    And just to avoid any misunderstandings if I had the funds I would be willing to get data on the notchbell, it’s interesting enough to merit it and get some solid data on its characteristics.

    I want to make it clear that anyone that wants to is welcome to play with anything I post. If it has value other than entertainment, good for you.

    I think mastery of controlling the flow separation is a bigger challenge for both ideas since both nozzles could easily experience big variations in stresses of the nozzle material.

    IMO the big problem is finding inexpensive ways to sort the wheat from the chaff. I am convinced that a compensating nozzle of some type can be done on the real cheap, it is just a matter of finding it and confirming it without financial distress.

  15. Tim says:

    I did mean a single chamber with three notches in the bell, although a multi-sided saddleback nozzle sounds interesting too. I was thinking you might be able to turn the structural weakness of the bell sections on its head. I was wondering about reinforcing the bell with struts, but then I realised that if the bell sections were flexible enough you might be able to attach servos to the bell that would deliberately distort them to acheive thrust vectoring. You might also be able to distort all three sections collectively to tighten control of the altitude compensation. I’m guessing self-locking servos would be a must to make this work, since you’d still need “struts” to hold the thing together.
    At the very least it would make an interesting exercise in FEA/CFD.

  16. johnhare johnhare says:

    Tim,

    I was wondering about reinforcing the bell with struts, but then I realised that if the bell sections were flexible enough you might be able to attach servos to the bell that would deliberately distort them to acheive thrust vectoring.

    I had never thought of flexible nozzle sections. If it could be made to work, it seems like it would make for a light compact package.

  17. Marcus Zottl says:

    Tim’s comment reminded me of thrust vectoring used for highly maneuverable fighter aircraft. Some concepts use few large flaps while others use a ring of many flexible elements to control the exhaust flow.

    Have those concepts been tried with rocket engines as well? To me it looks like such concepts could be used to achieve thrust vector control _and_ altitude compensation (I know this is the “wrong term” just can’t remember the proper one.. was it pressure difference compensation?) with one single concept.

    If you don’t know what I’m talking about, take a look at the pictures in the wikipedia article about thrust vectoring.

  18. Tim says:

    Marcus,
    That’s kind of where I got the idea from.

  19. john hare says:

    Marcus,

    Weight would be an issue with a straight adaptation of jet TVC systems. They need everything to work right for hundreds of hours and pay for it with mass. To use the concept though sounds feasible with Tims’ flexible nozzle concept. Relatively flat nozzle sections might bend in the proper manner without too much mass. Also they could possibly supply roll control with twisting the flaps. It seems possible to get three axis TVC with a single engine like that.

    Altitude Compensation is the proper industry accepted term. I am trying not to use it because it masks an important capability for VTVL vehicles when landing. Sometimes you want the ability to operate at high back pressure and low throttle which has less to do with altitude. Pc/pa compensating is more accurate technically, but not being in general use is even more misleading.

  20. Tim says:

    Just to add to the flexible nozzle thing, it might be possible to control a saddleback nozzle by flexing the endplates. I suspect this may be easier than flexing nozzle sections, since the endplates are naturally flatter.

  21. Another simple nozzle compensation option is the nozzle ring insert. Basically it’s like inserting an ablative solid rocket motor nozzle inside a conventional liquid rocket nozzle. The insert would be expelled at a desired staging altitude. According to this patent:
    http://www.freepatentsonline.com/5894723.html
    Two inserts for “two stages” could be used.

  22. john hare says:

    Roderick,

    While it didn’t strike me as being as simple as these two, it should be useful for expendables. Do you know if anybody has done any hardware testing of the concept?

  23. John:

    I haven’t found any mentions of actual tests, but that doesn’t mean they haven’t been done. I will lok to see if tests have been done.

    I would argue that it’s not just for expendables, even though this nozzle insert is, indeed, expended. For an RLV to “expend” something is no different than fighter aircraft unloading themselves of fuel drop tanks, or uisng a catapult to take off from an aircraft carrier, or the use of JATO rocket bottles on short runways.

    I concede, though, that this may not be simpler than the notched bell. However, wouldn’t nozzle flow be more problematic with a notched bell than with a nozzle insert?

  24. John,

    I can’t yet discern if these documents show actual testing, but they are related to the issue (both nozzle inserts and notched nozzles):

    http://www.patentstorm.us/patents/6574964/description.html

    http://www.wipo.int/pctdb/en/wo.jsp?IA=SE1998002222&DISPLAY=DESC

    An excerpt from the second link:
    “”””Attempts have been made to influence indirectly on the nozzle contour. Such prior attempts have included for example an exit diffuser means, see EP 626 513 A1, a trip ring, an abladeable or ejectable insert, periodically vari- able radius, see PCT/SE96/00176, slotted nozzles, see US-A- 5,450,720, injection of gas, see US 4,947,644, an exit ejector means, see EP 894 031 and a mechanism for extendi- ble exit nozzle, see EP 924 041. A flow separation control device with a modified nozzle contour is known from US-A- 3,394,549.””””””

  25. john hare says:

    Roderick,

    I’ll have to do some brainstorming on this. I don’t know for sure how much NIH prejudice I’m bring to the method.

Leave a Reply

Your email address will not be published. Required fields are marked *