From: ahahma@polaris.utu.fi (Arno Hahma)
Newsgroups: rec.pyrotechnics
Subject: Re: Composite Motors Part II: Propellants and Tampers
Message-ID: <1991Mar23.161910.20881@polaris.utu.fi>
Date: 23 Mar 91 16:19:10 GMT

In article <obuZDjy00Uh-E2dkVZ@andrew.cmu.edu> lc2b+@andrew.cmu.edu (Lawrence Curcio) writes:

>DRILLING A TAMPING DOWEL:

Deleted:
A long description how to make one with a press drill.

Use a lathe, it is much, much easier and faster. First drill the
hole through the dowel (which is best made of PVC or polyethylene),
if the dowel is long, use a cannon bore rather than a spiral one to get the
hole straight.
Next center the dowel between two spikes using the hole drilled
through. Turn the dowel round - it is ready. Now, the hole will be
exactly centered through the entire dowel.

>PROPELLANTS AND CASTING:

>with my limited technology was GALCIT-53. This is 80-85% KClO4 and
>15-20% asphalt by weight. Asphalt is available in hardware stores

>be kept more than a couple of months. They should be stored with
>the spintle (core former) inserted.

One of the problems with asphalt. Thiokol rubber would be much better.
It has disadvantages, the grain should be cast in a mold and cured in
an oven to make the water formed in the polymerization evaporate.
If the motor is small, casting directly into the motor and curing in
an oven works. With bigger motors, the time required for the water to
evaporate could be days, if the propellat was directly cast in. Also,
the oven must not be heated to over 100 C
immediately, but slowly. Otherwise the water starts boiling before the
rubber has cured and reached some strenght. This leads to fractures
and spaces within the fuel.

Thiokol also gives faster burning propellants, and the design of the
motor is easier - you need no long grains.

>would mix thoroughly. The mixture was then heated in an electric
>frying pan at 350 degrees F until the solvents were driven off and

>prevent debris falling into the propellant. Meanwhile, the engine,
>tamping dowels and wooden spoons were heated in an oven to the same

No heating required with polyurethane resins - easy and fast.

>the nozzle. I placed the tube on the nozzle form (I didn't trim the
>nozzle until after loading the propellant.) I had a dowel, drilled
>lengthwise as a tamper, which fit over the spintle. The propellant
>was packed a bit at a time, but the process was quick. IT WAS

The tubular grain you used is _very_ progressive. Considering the
high pressure exponent of the fuel, you really managed to make the
nozzle right. It has to erode with an exactly right speed to keep
the chamber pressure within the limits. In fact, this was difficult
to achieve, as I was testing with the shotgun shell rockets.
They also have a tubular channel due to the structure of the
shell. The nozzle material has to be stainless steel, because it melts at the
right speed! Anything more temperature resistant will cause a failure
(the grain flies out of the shell or the base is teared off). Anything
less resistant will cause a too low pressure and no thrust is generated.

>KClO4 instead of asphalt, but unless I cast the entire grain with
>a single push, I'd have a cracked grain explosions. (They are easy
>to spot - small, dull noise, and large chunks of engine/propellant
>lying around.) One-push grains must be short. They also waste a lot

Polyester does not work, because it is hard and brittle. A long
grain, no matter how well it is cast, will crack and explode when
fired. This is because a pressure gradient is formed between the
nozzle and the other end of the burning channel. The gradient will be
especially large in the tubular design, where the cross sectional
area of the burning channel is close to that of the nozzle throat.
This is also the case with nozzleless rockets - they require an
elastic fuel. Moreover, radial stress gradients build up in a thick-walled hard
tube, if the tube wall is not of exactly the same thickness and the support
on it (the glassfibre tube) is not exactly uniform. These
stress and pressure gradients will crack the grain. As a result,
a hard resin can not be used in case bonded grains.

An elastic resin can deform without cracking and turn the strain over
to the casing. It naturally can stand part of the stress itself.
This is the main reason, why polybutadiene has become a popular
binder - it is rubber-elastic. If new binders are developed, I am
quite sure, they will be rubber like as well.

The asphalt is rubber-elastic within a narrow temperature range. So,
it works in case bonded rockets, as long as the temperature is in
that range. This is why they are not used any more - the temperature
limits their use. An ideal binder has a low glass temperature and a high
melting point, i.e. a wide temperature range with elastic properties.

>required to prevent chuffing. The charcoal apparently aided in
>radiation transfer from flame to surface.

Moreover, it most likely prevented a resonant burning - tubular grains
are most prone to that, the longer the grain the easier it
resonates. REsonance can be avoided by adding inert materials to the
fuel or putting an inert rod into the combustion chamber of burning channel.

>composition, perhaps styrene/polyester and charcoal. This might
>help fuse the successive loads, or it might not. Also, it is

It would further help avoiding a resonant burning.

>possible that this rigid mixture, in a larger rocket motor, could
>be fractured by chamber pressure - particularly if it is more rigid

Or better, the joint between the inhibitor and the fuel would be easily
fractured, the flame would penetrate and increase the burning area
dramatically -> ....

>This mixture is not free flowing, and requires some light whacking
>with a mallet. It is rather like a pressed powder that is
>waterproofed with asphalt/rubber.

The loading closely resembles the loading of moist black powder.
A press is useful, you are able to cast the whole grain in one
batch, just use a form to support the casing during pressing.
With a form attached you can apply enough pressure to make the fuel
act as a fluid and flow evenly everywhere in the motor. Also, you need
no bored-through dowel - put enough fuel in to fill the motor over the
spintle top.

>This leads to erosive burning, but so what? Also, although cracks
>do form sometimes at loading interfaces, no explosions ever
>resulted. The burning rate was that low. Indeed, I tried a

And so is the minimal burning pressure. AN-propellants are good for
that reason, too. You need no strong casing. Also, the pressure exponent
of the AN-fuels is usually under 0,50, thus, they allow more nozzle
erosion and deviations in the burning area.

>CHUFFING: I.E; DISASTER!  - MAY HAPPEN ANYWAY!

Not, if the operating pressure is considerable higher than the minimal
burning pressure and the nozzle is plugged initially.

>horizontally and drops again. Successive bursts are longer, and
>they come with increasing frequency until stable burning results -
>all the while, your rocket is chasing you, everyone around you, and

>harmless. Chuffing is NOT. It is the VERY WORST THING THAT CAN
>HAPPEN ON THE PAD.

Yes, this is about the worst that can happen.

>ALSO USE COMPUTER SIMULATIONS! An engine that produces enough
>thrust to lift a rocket may nevertheless result in a powerdive in
>reality.

This is one of the reasons why short burning and high thrust engines are better
than low thrust long burning ones with the same impulse.

>have known parameters. ARNO has pointed out that the parameters
>will change according to particle size, however even a ballpark
>figure is better than no figure at all. Experimenting without doing
>the numbers is like hitting the beach with no air support. (See
>below!)

You are right. This is why building a strand burner is not a bad
idea at all - you are able to calculate the motor and it works
with the first try.

>I SHOULD ALSO POINT OUT!!!!! You can buy MUCH BETTER ENGINES than

May be, but at what cost?

>these. They have machined nozzles! They have more powerful
>propellants! They have gone though rigorous testing! They are even

Well, this applies to some self made engines, too....

>your choice. Let it be YOUR ASS, NOT MINE! The computer simulations
>still apply, however.

You are right. An error on the drawing board is much easier to correct than
an error in a rocket motor. I strongly recommend calculating the engines
before attempting to build one. Unfortunately, to be able to do this you need
the parameters of the fuel. This is the hard part of the whole process -
once you have the information, the rest is purely a problem of engineering.

______________________________________________________________________________

                              R'
                                \
                                 /======\
                               /          \        O
                             /     ArNO     \    //
                             \\        2    //--N+
                              \\          //     \
                                \\______//         O-
                               /
                             R

From: ahahma@kontu.utu.fi
Newsgroups: rec.pyrotechnics
Subject: Re: Progressive Grains, etc (was Composite Motors)
Message-ID: <1991Mar26.190536.37884@kontu.utu.fi>
Date: 26 Mar 91 19:05:36 GMT

In article <0buxoCu00WB60FcUVu@andrew.cmu.edu>, lc2b+@andrew.cmu.edu (Lawrence Curcio) writes:

> A cylindrically bored charge increases in burning surface
> linearly with the radius of the bore. My GALCIT bores started at

And exponentially in pressure, if the nozzle throat is constant in
area. If the web thickness of the grain is small, then the pressure perhaps
does not reach too high values.

> 5/16 inch and ended up as 12/16 inch. They were borderline low
> surface to begin with so, although some nozzle erosion was

Well, you had a smaller final area to initial area ratio than I had.
BTW, what percentage of those engines worked properly? I had a failure
percentage of about 20 with polyester based, case bonded engines, so I
abandoned polyester as a binder.

> I tried plugging my nozzles in various ways to hold in chamber
> pressure for stable burning. Results varied from explosions to
> nozzle blowouts, to propellant flameouts. I'd be interested in
> detailed schemes to overcome these problems. I'd  also appreciate

I calculate the plug to break above the lower combustion limit and below the
operating pressure. Since I choose the operating pressure to be at least
twice the lower limit, the bursting pressure of the plug is about 50 % of the
chamber pressure or somewhat more. The disc should be light or its inertia may
cause some additional rise of pressure, especially, if the ignition charge is
still burning, as the disc is released. This is often the case, at least with
black powder.

Of course, the nozzle has to be strong enough to hold the pressure for a while.
For example, if you have the conical section made of some light and not
so strong material, you should not glue the bursting disc on this part.
Instead, fasten it to the stronger parts of the nozzle, i.e. to the throat.

> details on strand-burner construction. It sounds ambitious but
> potentially worthwhile. Propellants that can be made to work in

Posting some drawings is somewhat difficult with ascii.. ;-). The principle is
simple, you burn a rod of propellant under constant pressure and measure
the burning time. Taking measurements at different pressures, you are able
to derive the burning equation of the fuel and calculate a motor.

Building a strand burner is not at all difficult, if you have some steel,
know how to start a lathe and what is a welding rod ;-).

The device is a pressure vessel equipped with a spring actuated valve, a
pressure gauge and about a dozen of electrical connections. The valve keeps
the pressure constant while the burning propellant generates more gas and heat.
The probes (to measure the burning speed) are simply thin copper wires running
through the strand. The vessel is initially pressurized with an inert gas
and the strand is ignited. I use the combustion gases of black
powder as the "inert gas" ;). I.e. the bomb is pressurized using a calculated
charge of black powder, which simultaneously ignites the strand. Time
between the melting of the test wires is measured giving the burning speed as
the distance between the wires is known.

This is only one possibility. If you make a window to the vessel, you can
track the flame front optically (it is really very sharp and bright) and
thus measure the speed. In this case, it is possible to measure all the
parameters in a single measurement. The burning acceleration, the pressure
and the derivative of the pressure are measured. From them, it is possible
to calculate the burning parameters. You can see this, if you differentiate
the burning equation. Measuring the burning acceleration (i.e. the derivative
of the burning speed) with the wire probe-method is less possible, you'd need
lots of wires close to each other to get enough measurement points.

The mechanical construction is not critical, just have enough space around the
strand to prevent any fast gas flows inside the vessel. I use a traditional
bolted flange to close the bomb, although it is not handy at all. However,
closing the bomb is the smallest job. Preparing the strand, insulating the
wires inside (against the high temperature) and the worst of all: cleaning
the vessel and the walls of the room you are using are a lot more work ;-).

> end-burning grains can have their parameters measured in
> miniature rockets with various nozzle diameters. Some black
> powder rocket companies actually do it this way.

It is also a more accurate way, the conditions are more realistic than in a
strand burner. The method has only one disadvantage: it is almost useless, if
you are measuring unknown propellants. If you have no idea, how fast the
propellant burns, then it is pretty hard to calculate a suitable nozzle for
the measurement. You would have to start with large diameters and slowly
change to smaller ones consuming perhaps tens of grains. Even if you know
the burning speed at some pressure, but have no idea of the pressure exponent,
you may have the same problem. If the pressure exponent is high, then a
small change in the nozzle may cause the propellant to extinguish or the motor
to explode. In addition, you need a very big collection of graphite nozzles
of a different diameters. The nozzle must be of graphite, any erosion will
destroy your results.

This is why I first use the strand burner, and then verify the
results using a test rocket motor (equipped with a pressure gauge) to get them
more accurate. It is simply easier and less work this way.

> Finally, it would be nice to see a description of the casting
> procedures for Thiokol, polybutadiene, and polyurethane. With the

All the propellants made of those resins are pourable or a dough, that can
be made to flow with a little help. The procedure is simple, mix the ingredients
, degass under vacuum (important!) and pour into the motor casing or a mold,
depending on what kind of a grain configuration you are using.
The percentage of the resin can be as low as 18 % with polyurethane and
polybutadiene and about 24 % with thiokol. Above these values the mass is
usually pourable. The pourability also depends on the oxidizer particle size and
distribution. A mix with all sizes of particles from 0 to some value (less than
5 mm) gives the best pourable mix. This is why they mix AP of different particle
sizes to the fuels. Surface active agents also help, adding a drop of Triton
X-**** helps a lot, if the fuel is too thick.

> generally available. Where can you get them, and in what
> quantities?

At least here it is easy. Just call the importer or manufacturer and order
some. The problem is, the smallest quantity is typically 50 kg or more and the
price in such a small lot is usually high. Various resins are widely used in
boat building, so, asking some stores often brings results. Thiokol
rubber, polyurethane and even hydroxy terminated polybutadiene are available
as boat resins. They are all used to glue the deck structures in sailboats.
It is a pity they are expensive.

ArNO
    2