Construction Classics

Accelerator Radiation Protection

Almost every shielding document in this collection was written for a machine far larger than anything an amateur will build, and the transfer problem is left to the reader. This one was written for the other end of the range. Its stated scope is the evaluation of design and operational safety for particle accelerators below 40–50 MeV, and its keyword list names cyclotrons and low energy explicitly. It is the only document here whose energy range starts where a small machine actually sits.

It also exists for an unusual reason, stated plainly in its foreword: in April 1969 a request to help the US Army Materiel Command draft an army regulation on particle accelerator radiation protection prompted a search for a reference that combined the vast amounts of information available in the literature, and none was found. So the Natick Laboratories staff wrote one. That origin is what makes it useful here: it is a document written to be the basis of a rule, by people who had to defend every number, rather than a laboratory reporting what it happened to measure.

Why this document can be hosted, and what was removed

The report is a work of the United States Army, prepared in-house by Natick Laboratories staff under an Army research programme. Works of the US Government carry no copyright. Its DD Form 1473 records it Unclassified, and its title page carries the distribution statement Approved for public release, Distribution unlimited. All 196 pages were searched for copyright notices, classification markings and distribution restrictions: none appear.

The acknowledgments, however, record one thing that a search for copyright notices would never have caught: Permission has been obtained for the use of copyrighted material from McGraw-Hill Book Company and The American Society of Mechanical Engineers. That permission ran to the Army. It does not transfer to anyone re-publishing the report, so the borrowed material had to be found before the document could be hosted.

Finding it took a page-by-page examination of all 196 page images, because the borrowed figures are identified nowhere except by superscript reference numbers in their captions, which the scan’s text layer drops. Five figures proved to be sourced from McGraw-Hill or ASME titles. All five were standalone figure plates, so this copy replaces those five pages with a notice naming the figure, its source and where to obtain the original. Nothing else is altered, no body text is affected, and the pagination is unchanged:

The chapters this site draws on most — II, V, VI and VII — are unaffected. The complete, unabridged report remains freely available from the DTIC copy at the Internet Archive, so nothing here is the only route to those five figures.

Three further McGraw-Hill citations were checked and are not reproductions: the Evans citation in Chapter VI supports a probability result quoted in running text; the Etherington handbook is attached to no figure or table, and the removal cross-section tables carry no attribution; and Chapter I’s accelerator tables carry no source note. One honest caveat: the acknowledgments are not a complete inventory of third-party material. Several Chapter IV figures are keyed to Physical Review and IEEE Transactions on Nuclear Science papers, which the acknowledgments never mention. Those captions read as describing the source experiment rather than reprinting its graphic, and no permission is claimed for them. Determination made August 2026, by reading the scan.

The eight chapters

Separately authored, and worth treating as eight short references rather than one book. Page numbers are the report’s printed numbering; the PDF runs nine pages ahead.

1. Accelerator Radiation Protection — Thomas G. Martin III · Chapter I, p. 1
   Scope and framing: what low-energy accelerators are used for, which committees set the limits, and how a radiation-protection programme is administered. The chapter that states the report’s energy range and its reason for existing.
2. X-Ray Production and Shielding — Ari Brynjolfsson & Thomas G. Martin · Chapter II, p. 13
   Bremsstrahlung from electrons stopping in matter: radiation length, forward intensity, angular distribution, spectra, and shielding thickness worked for concrete and lead with build-up factors and attenuation coefficients given as curves.
3. Neutron Production and Shielding — Francis J. Mahoney · Chapter III, p. 36
   Neutron production channel by channel — proton, deuteron, alpha and photoneutron — with threshold tables, yields per microampere against incident energy, and shielding by the removal cross-section method, including worked examples and a skyshine calculation.
4. Radioactivity Induced in Accelerator Facilities — Raymond D. Cooper, Thomas G. Martin & Ari Brynjolfsson · Chapter IV, p. 67
   What becomes radioactive and how much: activation cross-section tables, photonuclear yields by element, the elements that matter in shielding, targets, structure and cooling water, and expected dose rates from thick targets after one day, one week and one month of cooling.
5. Radiological Physics and Radiation Exposure Limits — Raymond D. Cooper, Francis J. Mahoney & Thomas G. Martin · Chapter V, p. 95
   The dosimetry underneath the limits: interaction processes, dose units, stopping power and LET, quality factors, kerma, biological effects, and the permissible-dose framework as it stood in 1972.
6. Measuring Radiation from Charged Particle Accelerators — Raymond D. Cooper & Francis J. Mahoney · Chapter VI, p. 112
   The chapter on pulsed-beam measurement, which is where accelerator health physics departs from reactor practice: instrument dead time, paralyzable and non-paralyzable counters, and the relationship between true and measured counting rates when the radiation arrives in microsecond bursts.
7. Access Limitations and Interlocks — Thomas G. Martin & Christopher W. Rees · Chapter VII, p. 135
   Physical barriers, shielding doors of nine types drawn to scale, labyrinth design worked geometrically, interlock logic, warning devices, emergency switches, and the radiation sensitivity of the electronic components the interlocks are built from.
8. Waste Effluents — Thomas G. Martin · Chapter VIII, p. 152
   Solid, liquid and airborne waste: transport groups and packaging rules under 49 CFR, activated cooling water, tritium monitoring, toxic gas production in irradiated air, and stack dispersion.

What transfers to a small machine

Chapter III is the one to read first if the question is whether a machine needs a neutron shield at all. It works production channel by channel with threshold tables and yield curves against incident energy, which is the form the question actually takes on a small machine: not how thick but is this channel open. The shielding deep dive makes that argument from other sources; this chapter is the same argument worked at length by a health physicist.

Chapter VI is the least duplicated content in the collection. Accelerator radiation arrives in microsecond bursts, so an instrument that reads correctly in a steady field under-reads badly in a pulsed one. The chapter derives the relationship between true and measured counting rate for paralyzable and non-paralyzable counters across three cases, and summarises them in one table. Anyone putting a survey meter next to a pulsed machine and trusting the number is solving this problem without knowing it.

Chapter VII is the practical one. Nine shielding door types drawn to scale, labyrinths worked as geometry with the scattering construction shown, interlock logic, and a section on how radiation-sensitive the interlock electronics themselves are — a failure mode that is easy to design straight past.

Read the energy range honestly. The report is scoped below 40–50 MeV, which is still well above a tabletop machine, and its worked examples are mostly electron accelerators and multi-MeV ion beams. What transfers is the method and the sequence: identify the channels your beam can open, compute a source term, attenuate, then verify by measurement. Its specific thicknesses do not transfer, and neither its numbers nor this page are a substitute for the current professional standard, NCRP Report No. 144, or for a qualified health-physics review.

About this copy

This is the Defense Technical Information Center microfiche-derived scan, which carries DTIC’s own notice that the copy furnished contained pages that do not reproduce legibly. It is honest about itself: the cover pages are noisy, and the text layer runs roughly 600 to 2,500 characters per page — usable for searching, unreliable for figures and equations. Read numbers from the page image, not from the text layer. The front matter alone garbles the report’s own numbering, rendering FL-169 as FL—163.

The five notice pages are set as real text rather than images, so they stay legible at any zoom and are findable by searching the PDF for removed. Page geometry is unchanged from the source scan; each notice takes the size of the page it replaces.

Design rules from this document

56 Design Guide rules come from this report, and they changed the shape of the guide rather than merely adding to it. Its two target domains had been the thinnest substantive ones here: shielding went from 12 rules to 40, and safety from 67 to 125. Shielding is no longer a stub, and this single document is why.

The distribution follows the chapters. Most are safety and shielding, with a cluster in detectors from Chapter VI — the pulsed-beam counting material, which had no real coverage here before. Each rule is attributed to its chapter author rather than to Martin, who edited; the 1972 exposure limits are carried as historical figures with the method rather than the numbers as the transferable part.

• dg-1033 — size the shield for the secondary radiation, not for the primary beam
• dg-1036 — bound the stray-electron X-ray source term when you lack design detail
• dg-1044 — neutron shielding is slow-down-then-capture, so concrete far outperforms lead
• dg-1047 — D(d,n) and T(d,n) are exoenergetic — any deuterium in the gas is a source term
• dg-1059 — a detector whose dead time exceeds the beam pulse under-reads a pulsed machine
• dg-1065 — the commonest cause of serious exposure is entry into the cell during operation
• dg-1070 — choose interlock components as carefully as the protection system itself
• dg-1074 — interlock bypass is inevitable, so pre-write the two-key procedure

Companion reading: ORNL-3540 chapter 11 is the design-stage shielding methodology from the accelerator side rather than the health-physics side, and residual radiation and personnel dose carries the measured activation data. Dose limits as regulation are at Legal; hazard identification is on Safety.