Maintaining and Advancing US Neutron Radiography Capability at the UC Davis McClellan Nuclear Research Center

Back of the Envelope Beginning

Back of the Envelope Beginning

Back of the Envelope Beginning

• If you want to detect 1 mm of water in between 1 cm slabs of aluminum which is better?
• X-rays: 56.9% vs 56.3% transmission is ~1% contrast.
• Neutrons: 81% vs. 63% transmission is a contrast of almost 30% or 30 times better than x-ray.

​

​

​

MNRC (1988)

F111 Airframe Life Extension Program (1990-1996)

​

​

MNRC Transfer from the Air Force to UC Davis in 2000

​

• The Air Force wanted to maintain MNRC’s capability but in an efficient way that they could not do it themselves.
• UC Davis was interested in acquiring MNRC to image titanium castings for DoD, silicon doping, boron neutron capture therapy, and I-125 production.
• All of these projects failed by 2002 and the center was considered the most troubled and under performing lab at UC Davis for the next 15 years, until they hired a new director in 2016.

A Little About Me

Neutron Radiography on Pyrotechnic Devices.

Neutron Radiography on Pyrotechnic Devices.

So Why Isn’t Everyone Doing Neutron Radiography?

Problem Statement

• In 2016 MNRC performed 1,500 radiographs while utilizing the reactor ~1,500 MWhrs.
• All staff from 2016 will be retired by 2025 (except for me) and industry has requested a capacity of 16,000 images per year.
• Due to our dwindling fuel supply MNRC is limited to 1,000 MWHrs of reactor utilization per year.
• And UC Davis was making me pay back the previous director’s 4-million-dollar debt.

​

Problem Statement

• The good news is that total neutron efficiency for the radiography process is about 0.000000001%.

​

Problem Statement

• The good news is that total neutron efficiency for the radiography process is about 0.000000001%.
• Making the target efficiency 0.000000015%.
• We modified radiography bays 3 and 4 to do commercial neutron radiography and staffed up from 6 staff members to 16.

​

So Why Isn’t Everyone Doing Neutron Radiography?

• Commercial radiography in the US is still film based, while research neutron radiography is overwhelmingly digital.
• Film-based neutron radiography requires a device known as a gadolinium conversion screen to sensitize the single sided x-ray film to neutrons.
• No one has successfully made gadolinium conversion screens since 1990.

So Why Isn’t Everyone Doing Neutron Radiography?

• Commercial radiography in the US is still film based, while research neutron radiography is overwhelmingly digital.
• Film-based neutron radiography requires a device known as a gadolinium conversion screen to sensitize the single sided x-ray film to neutrons.
• No one has successfully made gadolinium conversion screens since 1990.
• Until now…

2023 MNRC Fuel Inspection

Previous Inspection Criteria�

• Prior to November 2022, MNRC’s old license required annual inspection of only “lead” elements.
• These elements were defined as all elements (excluding IFEs) in C ring and surrounding the transient rod.
• This inspection methodology was based on the assumption that any issue with the core would be due to issues with the fuel from operating at high temperatures (based on 2.3 MW operation).
• This inspection methodology would not be effective at finding elements with manufacturing defects and issues with other core structures.

Element in A1

Failed Element (Positions D10)

• Insert pic of two elements

The Shim Plates

The Shim Plates

Elements in K1 and L1

The Upper Core Ring

Position L1

Potential Initiating Event

• In 1998 the fuel transfer cask “briefly rested” on the chimney structure and may have caused the core barrel or shim plates to shift. Documentation of the event is very limited.
• Observable design flaw in the shim plates/upper core ring and lower grid plate.
• Vibrations from >1.5 MW operation.
• Or a combination of these three potential causes.

​

Shim Plate Position Inspection Tool

The Results

• 2 of the shim plates were in the correct position.
• 2 of the shim plates had reduced clearance.
• 2 of the shim plates had no clearance, meaning fuel had been operated for 25 years at high power with no coolant on one side.
• 9 G-ring positions were eliminated, and 8 new elements were added to C-ring. Where did these new elements come from?
• Total downtime was 7 weeks, and the Nuclear Regulatory Commission issued us a license amendment in 4 days (which is a record). How did we get a license amendment issued so quickly?

How Did We Get a License Amendment Issued so Quickly?

How Did We Get a License Amendment Issued so Quickly?

The Gadolinium Conversion Screen

• Single sided x-ray film is only slightly sensitive to thermal neutron interactions.
• Gadolinium 157 has the largest thermal neutron capture cross section of any stable isotope of 255,000 barns with a natural abundance of 15.7%.
• 95% thermal neutron attenuation is achieved with by only 25 um of gadolinium.
• Gadolinium emits a number of conversion electrons after absorbing a thermal neutron.
• The conversion electrons have ranges of 50-100 um in water and less in gadolinium, sapphire, and film.
• These ranges match very well to the grain size and resolution of film.

Gadolinium Screen Interactions

Image taken from Neutron Imaging and Applications edited by Anderson, McGreevy, and Bilheux

The Gadolinium Conversion Screen

• Current screens are made on a 1/16^th of an inch 6061 aluminum substrate (14 inches by 17 inches) with 25 um gadolinium (metal) vapor deposited layer sealed by a 25 um sapphire protective lay also vapor deposited.
• Our current screens have held up well, but any flaws introduced into the screens cannot be fixed.
• With 14 screens at MNRC they have become a bottleneck.
• The are also the number one limitation to others doing commercial neutron radiography.
• The optical coating industry was not excited about the potential project.

Rethinking The Gadolinium Conversion Screen

• The screen must be gadolinium based and 14 by 17 inches.
• The critical dimension is not 25 um of gadolinium. A conversion screen needs only to be at least 25 um thick. Neutrons penetrating past 25 um essentially leave the system.
• The actual critical dimension is the protective layer.
• Too thick will completely attenuate the conversion electrons.
• A non-uniform thickness will non-uniformly attenuate the conversion electron resulting in a non-uniform exposure.
• Too thin of a layer may not sufficiently protect the gadolinium.
• No protective layer may not protect workers.
• After thinking about it for 2 years, why not just try a slab of gadolinium metal?

Making New Screens

Making New Screens

Making New Screens

• The metal plates were made to size. 14 inches by 17 inches by 1.0 mm.
• Graphite gadolinium inclusions had to machined off using a cutting wheel under flood coolant to machine off 20-30% of the plate’s thickness.
• Final sanding was performed to remove machining marks and any remaining inclusions.
• Finished plates were epoxied to a 1 mm thick aluminum backing to add rigidity.
• Finally, the protective layer was added.
• The end result was 14 new screens costing 1/3rd of the cost of used screens, repairable if damaged, 20% less image acquisition time, and still ASTM category 1.
• Current capacity at MNRC 20,000 images per year.

​

The Results

So Why Are We Trying So Hard?

So Why Are We Trying So Hard?

Questions?