space-rha.comShort courseChapter 2 · Shielding and transport

Chapter 2 · Lesson 1 of 6 · about 10 minutes

Radiation at the part

The radiation outside a spacecraft passes through structure, neighboring equipment and the package before it reaches a part. Along the way, particles slow down, change direction or create other particles. RHA needs the environment at the part, where the hardware actually operates.

Radiation at the part

Outside and inside

The environment model describes the field that arrives at the spacecraft. Transport works out how the surrounding material changes that field before it reaches the hardware.

Different RHA questions

Dose is the energy a particle field deposits in a stated material, accumulated over the exposure. Displacement damage and single-event effects are separate questions. Each needs its own exposure quantity and its own device-response data.

What a geometry model gives you

A mass model has a declared geometry, and that geometry supports path and mass calculations. A picture of the spacecraft or of the box does not by itself supply a flight radiation prediction.

Source and scoring boundaries

A transport request starts with the incoming radiation: the species, the spectrum, the angular distribution and the time interval. It then names the target: the location, the material and the quantity to calculate. A field at the box wall, a field at the die and an absorbed dose are three different products, and the request should say which one you want.

Source definition. Record the species, the energy variable, whether the spectrum is differential or integral, the units, the angular distribution, the boundary on which the source is defined, the time interval and the statistical case the environment model was run for. Keep the source populations separate, so that each one's contribution to the result can still be identified later.

Target definition. Give the target's coordinates and orientation, its material, and the volume or surface over which the score is taken. Say which quantity you want scored: the fluence crossing a surface, the local spectrum, the deposited energy or the dose in a named material.

Package responsibility. Say whether the package, the lid, the die overlayer and the board are inside the transport model or outside it. The response calculation that follows has to use the same boundary, otherwise the same material can be counted twice.

Time cases. A mission-lifetime integral, the exposure from a single event and a short operating window are three different requests, even for the same part in the same box. The geometry does not change between them, so one geometry model can serve several source cases as long as each source case is under configuration control.

The local LET spectrum. The precomputed integral LET flux examples in the reading list are for silicon behind 100 mil Al. At that fixed shielding you can compare orbits or locations, read off F(>L), or difference the curve into finite LET bins. A local LET spectrum is still only an exposure. Before it becomes an SEE rate it needs a device response and an angular convention that match it.

3D scene

From the external field to the part

The boundaries that transport crosses: the environment outside, the structure and box, the package and the die.

  1. The environment model describes the field that arrives at the spacecraft: the species, the spectrum, the angular distribution and the time interval, on a stated boundary and for a stated statistical case. The teaching enclosure is shown assembled; nothing about the inside is known yet.

  2. Opening the box shows the structure, the board and the package around the target, and the rays sample the material in every direction. Transport works out how that material changes the field along each of them. The request has to say whether the package, the lid, the die overlayer and the board are inside the transport model or outside it, because the response calculation that follows has to use the same boundary.

  3. The product is a named quantity at a named location: the fluence crossing a surface, the local spectrum, the deposited energy or the dose in a stated material. A field at the box wall, a field at the die and an absorbed dose are three different products, and a mission-lifetime integral, a single event and a short operating window are three different requests for the same part in the same box.

The original space-rha.com environment and mission renderer (NASA imagery and spacecraft models, credits in the view), with the course scenes drawn by the same renderer. Trails, clocks and counts are illustrative; the scene computes no dose or rate.

Exercise

Write a transport request that another engineer could run without coming back to you with questions: the source and the target defined in compatible terms.

Write down the source, the geometry and the requested score before you choose a transport tool, so that the tool is chosen to fit the question.

Going deeper

Angles, areas and normalization

An angular intensity, an omnidirectional flux and a count of surface crossings are measured differently. A Monte Carlo score is reported per generated history. It becomes a physical exposure only after the source sampling, the particle weights, the area, the solid angle and the duration have been reconciled with the physical source.

Angular measure. For an isotropic directional intensity j, defined per unit area normal to the direction of travel, the integral over the full sphere is 4πj. The rate at which particles cross a flat surface from one side, per unit area of that surface, is πj, because the projected area brings in a cosine. The two are different physical quantities.

Source sampling. The source generator may sample from a sphere, a plane, a volume or a biased region of phase space. Its probability density and weights have to represent the physical source you specified. Do not apply a 4π or source-area factor from habit; derive it for the generator you actually used.

Deposited energy. Absorbed dose is the deposited energy divided by the mass of the scoring volume, with 1 Gy = 1 J/kg. The score has to say which secondaries were followed and in which material the energy was deposited. A score per primary particle becomes a physical dose only through the number of physical primaries, so that normalization has to match the source definition.

Boundary closure. A surface-crossing tally may count entries, exits, repeated crossings or only selected directions. A track-length fluence estimator is a different quantity again. Write down which estimator and which boundary rule each code used before you compare their results.

Takeaway

Start by naming the location of the hardware and the effect you have to assess.

Check yourself

A transport result is reported as a flux per cm² per second. What else do you need to know before you can use it?

Whether it is omnidirectional or per steradian, integral or differential in energy, at what depth or interface it was scored, and in which material the LET or dose is expressed. Without those the number cannot be compared with a test result or a requirement.

Why is the dose at the part generally different from the dose behind a slab of the same thickness?

The part sits inside a structure that shields it differently in every direction, and the particles arrive from all directions. The slab value is one direction and one thickness; the part value is the average over the solid angle, weighted by what each direction lets through.

A colleague sends a detailed rendering of the box and asks for the dose at the part from it. Why is that not enough, and what does the transport request need instead?

A picture of the spacecraft or of the box does not by itself supply a flight radiation prediction. Transport needs a mass model with a declared geometry that supports path and mass calculations, a target with its coordinates, orientation, material and the volume or surface over which the score is taken, and a source stated by species, spectrum, angular distribution, boundary, time interval and statistical case. The request should also say which product is wanted, since a field at the box wall, a field at the die and an absorbed dose are three different things.

The transport model includes the package and the lid, and the SEE rate calculation that follows uses a response model that also carries the package. What goes wrong?

The same material is counted twice, once in the transport and once in the response. The request has to say whether the package, the lid, the die overlayer and the board are inside the transport model or outside it, and the response calculation that follows has to use the same boundary. Lesson 2.6 makes the same point from the other side: the package is transported once, either in the local-field stage or in the response stage, and not in both.

An isotropic field is given as a directional intensity j. A colleague multiplies by 4π to get the rate at which particles cross a flat detector surface from one side. What is the error?

The integral of j over the full sphere is 4πj. The rate at which particles cross a flat surface from one side, per unit area of that surface, is πj, because the projected area brings in a cosine, and the two are different physical quantities. The same caution applies to a Monte Carlo score, which is reported per generated history: the 4π or source-area factor should be derived for the generator actually used rather than applied from habit.

Further reading