Ignition Transient

From BurnSim Wiki
Jump to navigation Jump to search

The ignition transient is the start of a solid rocket motor's burn, while the chamber fills from ambient up to operating pressure. BurnSim's Ignition tab models that fill and reports how long it takes: the 10, 50 and 90% pressure rise times, when the nozzle chokes, and the chamber's L*. You set how much of the propellant is alight at the start and how fast the flame spreads, or let BurnSim fit both to a measured firing.

Commercial license only, and experimental. The tab sits beside Results, Test Data, Notes and Snapshots in the bottom-right panel, labeled Ignition (experimental). It is hidden on personal, educational and trial licenses. See the purchase page.

What it does

The main simulation is quasi-steady: at every step the chamber is at the pressure where the gas the propellant makes equals the gas the nozzle passes, so a motor is at full pressure from the first time step. That is a good model of the burn, and wrong for its first few milliseconds, when the chamber is still full of air and the nozzle may not even be choked yet.

The Ignition tab integrates that fill as a separate calculation. It starts the chamber at ambient pressure, lights the propellant you say is lit, lets the flame front spread over the rest, runs the nozzle unchoked until the pressure ratio crosses its critical value, and follows the pressure up to the same equilibrium the Results tab reports. It changes nothing about the main thrust curve, burn time or total impulse: those are exactly what they are without it.

Why it matters:

  • Igniter work. How quickly a motor comes up to pressure, and how that changes with how much surface the igniter lights, is the question an igniter is designed around.
  • Staged or partial ignition. A long grain stack lit at one end, or a single grain lit over part of its surface, comes up in steps rather than at once. The tab shows the shape.
  • Free volume. The fill time is set by the chamber's L*, so the forward and aft ullage you leave in the case directly sets how slow the start is.
  • Test data. Comparing a predicted rise against a measured one, with the measured trace drawn on the same axes, tells you how much of a slow start is the chamber and how much is the ignition.

Inputs

All four are saved with the motor.

Grains lit at t=0
How many grains, counting from the forward (head) end, are alight at the instant ignition starts. 0 means the whole stack lights together, which is also how every motor saved before this field existed is simulated. Unlit grains make no gas until the flame front reaches them.
Or surface lit at t=0
The percent of the whole propellant surface alight at t=0, spreading to the rest as the flame front travels. Anything above 0 overrides the grain count. This is for the case the grain count cannot describe: a single grain lit over part of its surface.
Flame speed
How fast the flame front travels along the stack, in in/s (m/s in metric). Used only when part of the motor is unlit at t=0, and grayed out otherwise. Left at 0 with something unlit, it is filled in at 60 in/s, a starting point taken from two measured firings and not a property of your motor. A speed, rather than a time per grain, so a value fitted on one motor carries over to a longer one.
Ignition delay
Milliseconds from the igniter firing to the propellant surface lighting. It only shifts the curve along the time axis, so the prediction can be lined up with a trace whose t=0 is the moment the igniter was fired.

The ignited area and the flame speed are values you supply, not predictions. A lumped (0-D) chamber model cannot compute flame spreading; this one lights whatever you tell it to light. The honest source for both is a measured firing, which is what Fit to measured is for.

Below the inputs:

  • Show measured draws the test-data set selected on the Test Data Tab on the same axes, at its own t=0.
  • Fit to measured finds the surface lit at t=0, the flame speed and (with a fire-marker trace) the ignition delay that make the predicted rise match the measured one, and fills them in. See below.
  • Calculate runs the fill for the current motor and ignition condition.

What it reports

Each figure is shown with its definition beside it, beside a plot of chamber pressure against time.

Ignition interval
The time for the whole 0-90% rise.
Milestones
The times at which the chamber reaches 10%, 50% and 90% of the rise.
Equilibrium
The pressure the chamber settles at. It is the same figure the Results Tab reports, reached by integrating rather than assumed.
Nozzle chokes
When the flow through the throat first goes sonic. Read it loosely: it happens while the chamber still holds mostly air.
Initial L*
Free volume divided by throat area, which is what sets the timescale of the fill.
Flame speed
Shown when part of the motor starts unlit, with a note if it was filled in for you.

The milestones are fractions of the rise above ambient, not of the absolute chamber pressure. That is how they are read off a measured trace, which starts at ambient rather than at zero. On a high-pressure motor the difference is small. On a low-pressure one it is not: on a 145 psi motor, ambient is already 10% of the absolute pressure, so a 10% milestone measured from zero would read as time zero.

L* and ullage

L* ("L-star", the characteristic chamber length) is the chamber's free volume divided by its throat area. For a given propellant, the time the chamber takes to fill scales in proportion to it: double the L* and the fill takes about twice as long.

The free volume is the core, plus the forward and aft ullage - the open space ahead of the first grain and behind the last one. Both are entered as lengths on the Motor tab (Fwd ullage and Aft ullage, see BurnSim Fields) and drawn in the cross-section view. They default to 0. A motor with a real gap at either end that leaves them at 0 will show a faster rise than it has. The ullage fields and the Initial L* line on the Results tab come with every license; only the Ignition tab itself is commercial.

Fitting to a measured firing

Select a firing on the Test Data Tab, tick Show measured, and click Fit to measured. BurnSim searches the lit percentage, the flame speed and, when the trace has a fire marker, the ignition delay, and fills in the best match.

It fits the shape of the rise, not its level. If the measured motor settles at a different pressure from the predicted one, that is a burn rate or C* question, not an ignition one; the fit reports the mismatch rather than bending the ignition inputs to hide it. Fix the level with Propellant Characterization first.

From a single trace the flame speed and the delay are pinned well and the lit percentage only loosely. A fire marker - a spare data channel recording the moment the igniter fires, see Data Acquisition - is what lets the delay be measured at all; without one, the trace's t=0 is the start of the pressure rise and there is no delay to find.

What it does not model

This section is the important one. Read it before comparing a prediction with a test.

  • Lighting the propellant is usually the slow part, and the ignited area is your input. A small motor's chamber fills in a few milliseconds once its surface is burning, but the surface itself can take tens of milliseconds to light. On one measured 54 mm single-grain firing, the whole-surface-lit prediction rose in about 7 ms and the motor took about 38. A prediction with everything lit at t=0 will come up much faster than a real motor.
  • The igniter's own gas is not modeled. On its own that makes the predicted rise slower than reality, most on a small motor with a gassy igniter. It pulls the opposite way from the point above, and which wins depends on the motor and the igniter. Read the times as how fast this chamber fills for the ignition you entered, not as a prediction of your trace.
  • No igniter, so no overshoot. BurnSim does not currently predict a pressure overshoot at ignition. Without an igniter term the model cannot produce one, so the tab does not show an overshoot figure rather than show one that always reads zero. The same goes for induction delay and hangfire margin: they need an igniter model.
  • No pressure waves. The spike that can come from pressure waves traveling down a long, narrow port is one-dimensional, and a lumped chamber has one pressure everywhere.

The tab is marked experimental because it is still being checked against measured firings, and it says so in a banner across the top. Treat its times as indicative, not as a number to design to.

See also