Design Innovationdesign Innovation Papers

8/13/2019 Design Innovationdesign Innovation Papers

1/21

Design InnovationDESIGN INNOVATION PAPERS

Design of a Magnetic Resonance Imaging Compatible Metallic Pressure Vessel

Matthew Ouellette,Hui Han,Bryce MacMillan,Frederic Goora,Rodney MacGregor,Marwan Hassan and

Bruce J. Balcom

[+-] Author Affiliations

Matthew Ouellette

MRI Research Centre,

Department of Physics,

Mechanical Engineering Department,

University of New Brunswick,

Fredericton, New Brunswick, E3B 5A3, Canada

Rodney MacGregor

MRI Research Centre,

Department of Physics,

University of New Brunswick,

Fredericton, New Brunswick, E3B 5A3, Canada

Marwan Hassan

Mechanical Engineering Department,

University of New Brunswick,

Fredericton, New Brunswick, E3B 5A3, Canada

Bruce J. Balcom

MRI Research Centre,

Department of Physics,

University of New Brunswick,

Fredericton, New Brunswick, E3B 5A3, Canada

e-mail: [email protected]

Corresponding author.

Contributed by the Pressure Vessel and Piping Division of ASME for publication in the Journal of Pressure

Vessel Technology. Manuscript received March 27, 2012; final manuscript received February 4, 2013;

published online June 11, 2013. Assoc. Editor: Allen C. Smith.

J. Pressure Vessel Technol. 135(4), 045001 (Jun 11, 2013) (7 pages)doi:10.1115/1.4023728History:

Received March 27, 2012; Revised February 04, 2013

8/13/2019 Design Innovationdesign Innovation Papers

2/21

Article

References

Figures

Tables

Abstract

Abstract |Introduction |MRI Background |MRI-Compatible Metallic Pressure Vessel Design |Design

and Constraints |Imaging Results |Conclusion |Acknowledgements |References

High-pressure measurements in most scientific fields rely on metal vessels, a consequence of the

superior tensile strength of metals. Magnetic resonance imaging in conjunction with metallic pressure

vessels has recently been introduced. Magnetic resonance imaging with compatible metallic pressure

vessels is a very general concept. This paper outlines the specifics of the development and design of

these vessels. Metallic pressure vessels not only provide inherently high tensile strengths and efficient

temperature control, they also permit optimization of the radio-frequency probe sensitivity. The design

and application of magnetic resonance imaging compatible pressure vessels is illustrated through a rock

core holder fabricated using nonmagnetic stainless steel. Water flooding through a porous rock at

elevated pressure and temperature is shown as an example of its applications. High-pressure magnetic

resonance plays an indispensable role in several scientific fields; this work will open new avenues of

investigation for high-pressure material science magnetic resonance imaging.

FIGURES IN THIS ARTICLE

8/13/2019 Design Innovationdesign Innovation Papers

3/21

8/13/2019 Design Innovationdesign Innovation Papers

4/21

8/13/2019 Design Innovationdesign Innovation Papers

5/21

8/13/2019 Design Innovationdesign Innovation Papers

6/21

8/13/2019 Design Innovationdesign Innovation Papers

7/21

8/13/2019 Design Innovationdesign Innovation Papers

8/21

8/13/2019 Design Innovationdesign Innovation Papers

9/21

Introduction

Abstract |Introduction |MRI Background |MRI-Compatible Metallic Pressure Vessel Design |Design

and Constraints |Imaging Results |Conclusion |Acknowledgements |References

Over the past half century high-pressure magnetic resonance (MR) has played an important role in many

areas of physics and chemistry [1-6]. Specifically MR measurements at high pressure and variable

temperature provide unique information about the microscopic behavior of liquids [2,3]. Magnetic

resonance imaging (MRI) has revolutionized clinical diagnostic imaging and its use exploring and

understanding other natural science systems is expanding. There is clear merit in combining MRI with

high pressure and temperature control for materials investigation.

We have previously introduced high-pressure magnetic resonance imaging using metallic vessels from a

magnetic resonance point of view [7,8]. This article describes the process of designing these vessels, the

ultimate design goals, and their effects on the design and fabrication. The material selections, sizing

(diameter, wall thickness, length), sealing mechanisms, and radio-frequency (RF) probe design are all

governed by the design goals and unique limitations imposed by MRI. For a pressure vessel the mostnotable design goals are the design pressure and temperature range.

The metal vessel encloses a pressurized RF probe with the metal enclosure functioning as an RF shield.

The naturally high thermal conductivity of metals allows the sample space temperature to be monitored

and controlled by regulating the outside temperature of the vessel. The metallic pressure vessels

imagined are very general in concept; they may be closed vessels of various designs including chemical

reaction vessels, high-pressure pipes, or rock core holders with radial pressure application.

In this article a rock core holder is used as a specific example of a metallic pressure vessel. A core holder

is employed to maintain a cylindrical rock sample at high pressure and specified temperature in order to

recreate petroleum reservoir conditions. The design pressure and temperature for this core holder were

chosen to be common, but not extreme, conditions of 6.9 MPa (1000 psi) and 50 C. Once a rock core

sample is at reservoir conditions, MRI can be employed to analyze the fluid behavior to obtain

information about the nature of the reservoir. As an illustration of this utility, water flooding a porous

rock under pressure and temperature control is monitored by MRI.

8/13/2019 Design Innovationdesign Innovation Papers

10/21

A limited number of high-pressure MRI studies have previously been reported [9-12]. These studies

almost universally employ nonmetallic polymer composite cells that fit into an existing MRI RF probe.

These measurements are hindered by low sensitivity of the RF probe due to the distance between the

probe and sample under investigation [9-11] as well as inefficient temperature control due to low

thermal conductivity of the polymer [9-11].

MRI Background

Abstract |Introduction |MRI Background |MRI-Compatible Metallic Pressure Vessel Design |Design

and Constraints |Imaging Results |Conclusion |Acknowledgements |References

Magnetic resonance occurs for atomic nuclei having odd numbers of protons or neutrons, and thus

possessing a net magnetic moment. Hydrogen is the most commonly examined nuclei; MRI is most

sensitive to hydrogen, and most desirably H2O and hydrocarbons can be investigated. In the presence of

a static magnetic field (B0), the magnetic moments are polarized resulting in a net magnetization vector

(M0) in the B0 direction. The nuclear spins can be excited with a radio-frequency (RF) field (B1) applied

perpendicular to B0 at the Larmor frequency given by [13]:

(1)f 0 =B 0

Where is the gyromagnetic ratio and B0 is the magnetic field strength. The gyromagnetic ratio is a

constant and depends on the nuclei being examined, in the case of 1H it is 42.58 MHz/T. RF excitation

causes M0 to depart from equilibrium and precess about the static field. Precessing magnetization

induces an electrical potential difference in the RF coil when switched to reception mode. The system

will return to equilibrium with characteristic MRI signal relaxation times T1, T2, and T2*. In the case of a

centric scan SPRITE MRI method, commonly employed for quantitative studies of fluids in porous media

[14], the local image intensity is given by

(2)S=M 0 exp(t pT 2 )sin

Where tp is the phase encoding time, and is the RF rotation angle. The net magnetization vector M0 is

proportional to quantity of nuclei; therefore the induced voltage in the RF probe will be directly

proportional to local fluid quantity 0.

(3)S 0 exp(t p T 2 )sin

Switched magnetic field gradients are employed to spatially resolve the signal in one, two, or three

dimensions [15,16]. The signal intensity at each pixel or voxel will be governed by Eq.(3);fluid quantity

and relaxation time information is measured and spatially resolved.

MRI-Compatible Metallic Pressure Vessel Design

Abstract |Introduction |MRI Background |MRI-Compatible Metallic Pressure Vessel Design |Design

and Constraints |Imaging Results |Conclusion |Acknowledgements |References

8/13/2019 Design Innovationdesign Innovation Papers

11/21

Figure 1(a) shows a schematic of a general MRI compatible metallic pressure vessel. The metallic vessel

encloses a pressurized MRI RF probe with the metal enclosure functioning as an RF shield. The metallic

pressure vessel, surrounded by a temperature control coil, fits inside a standard gradient bore on an

MRI spectrometer. Figure 1(b) shows a conventional nonmetallic vessel. The RF coil is placed outside of

the vessel.

Pressure Vessel Stresses.

Conventional (non-MRI) laboratory studies at high pressure rely on metal vessels given their superior

tensile strength. The high tensile strength of metals permits larger macroscopic volumes to be studied

by MRI. As the size of a pressure cell increases, the required yield strength of the cell must increase

proportionally to maintain the same maximum internal pressure. For a thick-walled cylindrical pressure

vessel the radial stress r and circumferential stress vary as a function of radius r as *17+

(4) r =P(r 2 i r 2 o r 2 i )(r 2 i r 2 1)

(5) =P(r 2 i r 2 o r 2 i )(r 2 i r 2 1)

where P is the pressure difference from ambient pressure, ri is the inner radius of the cylinder, and ro

is the outside radius. Longitudinal stress z in the cylinder is constant and described as

(6) z =P(r 2 i r 2 o r 2 i )

The maximum allowable stress in the cylinder will be a yield criterion stress divided by a safety factor SF.

For this design a conservative safety factor of three was assumed. The most common yield criterion for

ductile materials is the Von Mises stress VM. The Von Mises stress assumes that failure occurs when

the energy of distortion reaches the same energy for yield/failure in uniaxial tension. Mathematically

this is expressed as

(7) VM =1 2 *( r z ) 2 ( r ) 2 ( z ) 2 +

The maximum Von Mises stress will occur on the inside wall of the cyldinder where r = ri.

The high yield strength of stainless steel makes it a near universal choice for conventional pressure

vessels. A variety of metals may be employed for fabricating the cylindrical vessel including aluminum,

nonmagnetic stainless steel alloys, brass, and titanium. The physical properties relevant to MRI

compatible pressure vessels are shown in Table 1 [18,19]. Stainless steel has a very high tensile strength,

400 MPa *18+, twice that of aluminum.

For a nonmetallic MRI-compatible pressure vessel thicker walls are required because of the inherently

lower tensile strength of the materials. For a constant inside radius (sample space) and design pressure,

the wall thickness required (ro ri) will change, depending on the strength of material used. The wall

thickness required for an aluminum vessel is roughly twice as large as that of a vessel made of stainless

8/13/2019 Design Innovationdesign Innovation Papers

12/21

steel, and vessel made of polyether ether ketone (PEEK) is nearly four times the thickness of a stainless

steel vessel.

Temperature Control.

Operation at variable pressure and temperature are critical requirements for pressure vessels.

Temperature control is consistently problematic in MRI since MR observable fluids (notably water)

cannot be readily employed. The water will superimpose as signal over the sample being measured,

where a metal vessel will shield this signal. Commercially available solutions, such as those available

from Temco [11,12] involve circulating fluid under high pressure through the pressure vessel. The

confining fluid must be temperature controlled, pressurized, and circulated. This is a complicated and

expensive solution because of the specialized pumps and equipment required. One may however readily

exploit the high thermal conductivity of metals, forming the wall of a metallic pressure vessel, for a

much simpler and more passive method of temperature control.

The high thermal conductivity of metals as outlined in Table 1, allow the sample temperature inside the

vessel to be manipulated by controlling the outside temperature of the vessel and letting the system

reach equilibrium. The metallic vessel can be modeled as a hollow cylinder, with the instantaneous rate

of heat transfer per unit length into the cylinder Q (W/m) a function of inside and outside radius ri and

ro, inside and outside temperatures Ti and To, and thermal conductivity k as follows [20]:

(8)Q=2k(T i T o ) ln(r o /r i )

If the outside temperature To is held constant, heat will be deposited into the sample space

temperature Ti, at rate Q. The rate of heat transfer will decay in proportion to the temperature gradient

shrinking as inside temperature approaches the outside. At steady state the heat equation for a hollow

cylinder is given as

(9)d dr (rdT dr )=0

Assuming a constant outside temperature, the general solution for temperature as a function of radius

at steady state is

(10)T(r)=Qr i k ln(r o r )+T o

Substituting Eqs.(8) into(10) shows that at steady state (after a long enough time period in this case),

the outside temperature will be equal to the inside. The high thermal conductivity of metals allows for a

high rate of heat transfer; therefore the sample approaches steady state relatively quickly. This method

is not possible with conventional MRI-compatible pressure vessels made of polymers with low thermal

conductivities.

The temperature control setup for the metallic core holder is much simpler compared to commercial

solutions. Applying nitrogen gas from a gas cylinder pressurizes the sample space. The regulation of the

outside temperature is achieved by pumping temperature-regulated water through thermally

conductive tubing in contact with the exterior of the vessel. Water is a natural choice given the high

8/13/2019 Design Innovationdesign Innovation Papers

13/21

thermal capacity of water and ease of handling water flow. Unlike nonmetallic pressure vessels, one

need not be concerned about MRI detection of the external water by the RF probe since the metallic

vessel acts as an RF shield. This temperature control strategy is simple, robust and advantageous

because it separates temperature control from pressure regulation.

The Nitronic 60 nonmagnetic stainless steel core holder was tested for temperature control. Thetemperature was monitored with a thermocouple attached inside the sample space. As expected,

temperature changed exponentially with time. To reach a sample space temperature of 50 C from room

temperature required an hour.

Probe Sensitivity.

MRI is based on the manipulation and detection of nuclear magnetism, and nuclear magnetism is a

relatively weak physical effect. This leads to MRI being a relatively insensitive measurement technique,

but the metallic vessel design offers two unique benefits that help greatly increase the sensitivity of the

MRI experiment.

Placement of the RF coil in close proximity to the sample space maximizes the probe sensitivity, and the

metallic vessel acts as an electromagnetic interference (EMI) shield for the RF probe reducing

background interference and noise. In previous work the background noise level was measured for the

probe without any RF shielding, the probe enclosed within the steel core holder, and the probe enclosed

in a typical copper shield. The background noise for the metallic vessel was found to be an order of

magnitude less than the unshielded case [7]. The background noise for the conventional copper shield

was roughly twice the metallic vessel [7].

The principle of reciprocity suggests that maximizing the RF coil sensitivity requires maximizing the RF

magnetic field B1 per unit current in the sample space [21]. The B1 field (at the center of the coil) perunit current is inversely proportional to the characteristic size of the probe coil [22]. For example the

magnetic field B at the center of a current loop for a circular coil with diameter D is

where 0 denotes the vacuum magnetic permeability and I is the total current in the loop.

The ultimate sensitivity S of an MRI experiment at sample temperature Ts and RF probe temperature

TRF is described as [22]

(12)SB 1 I 1 T s T RF

From Eq.(11),the ratio of B1/I is inversely proportional to the characteristic size of the RF coil, which

when combined with Eq.(12) gives the relationship

(13)S1 D 1 T s T RF

The value of 1/D is at least two times larger for a metallic vessel than for nonmetallic vessel and more

likely a factor of 4 because the probe is inside the vessel, not outside.

8/13/2019 Design Innovationdesign Innovation Papers

14/21

One must consider that the RF probe in a metallic vessel will be maintained at the sample temperature

Ts instead of the magnet bore temperature Tmag and as a consequence will have increased thermal

noise as the sample temperature is increased as suggested by Eq.(13).This is fundamental to the design

and a potential drawback compared to the conventional approach. Calculating the sensitivity to

compare the effects in the metallic vessel (favorable 1/D, less favorable TRF = Ts) to those in the

nonmetallic one (less favorable 1/D, favorable TRF = Tmag) show the metallic vessel to be highly

favorable. Decreasing the sample temperature below ambient lab temperature improves the situation

where TRF = Ts. The thermal noise does not balance the effects of the increased sensitivity of the RF

probe until the sample temperature reaches 3600 C. At the design temperature of 50 C, the ultimate

sensitivity is 3.8 times greater for the metallic vessel versus the nonmetallic case. Both of these results

prove the RF probe for a metallic vessel to be very sensitive and therefore very effective.

Eddy Current Effects.

Switched magnetic field gradients are essential to spatial encoding and motion sensitization in modern

MRI. As the magnetic field gradient switches, eddy currents will circulate in metals and may distort the

desired magnetic field in the sample space. We have previously examined the eddy current problem

associated with MR imaging inside conductive metallic structures [8], and how it may be solved by

magnetic field gradient measurement and data correction. Eddy currents induced in a cylindrical metal

vessel decay exponentially, with a time constant dominated by the following term [8,23]:

(14)=K

where and are the electrical conductivity and magnetic permeability of the material, K is a constant,

and represents a geometric factor. Illustrated in Table 1, a stainless steel vessel compared to aluminum

reduces electrical conductivity and therefore eddy current time constants by a factor of almost 30

[18,19].

Gradient waveforms were simulated for both Nitronic 60 stainless steel and aluminum cells using CST

EM Studio [24]. Appropriate geometries were selected for comparison with magnetic field gradient

waveform measurements [25]. The simulations showed aluminum had a severe eddy current problem

while the Nitronic 60 stainless steel did not. In the case of Nitronic 60 the eddy current effect decayed

within 0.85 ms, while for aluminum it lasted 32 ms. Simulation and experiments agreed very well; the

difference between gradient rise times measured experimentally and determined through simulation

was less than 5%.

Design and Constraints

Abstract |Introduction |MRI Background |MRI-Compatible Metallic Pressure Vessel Design |Design

and Constraints |Imaging Results |Conclusion |Acknowledgements |References

A core holder was chosen as an application of the MRI compatible metallic pressure vessel concept. A

core holder is employed to maintain a cylindrical rock core sample at high pressure and specified

temperature in order to recreate petroleum reservoir conditions. A core holder applies a static radial

8/13/2019 Design Innovationdesign Innovation Papers

15/21

pressure and variable axial pressure to force a flow in the longitudinal direction for measurements. In

this case a design pressure and temperature of 6.9 MPa (1000 psi) and 50 C was chosen, with a safety

factor of three chosen for the design. Figure 2(a) depicts a core holder embodying the principles

outlined in Fig. 1(a). The metallic vessel wall encloses an RF probe that in turn contains a sample. Access

through the end plugs is provided through threaded holes for fluid inflow/outflow and electrical

connection to the RF probe.

The type of RF probe employed will vary depending on the style of magnet employed. In the case of

superconducting magnets, a birdcage or saddle coil is used [22]. Figure 2(b) shows the construction of

the probe. The RF probe is constructed of two concentric hollow cylinders with the RF coil embedded in

the diametrical clearance between them. The RF probe assembly is sealed with a high strength epoxy

impermeable to water. A conventional RF shield is not required since the vessel wall performs this

function.

Figure 2(c) shows a photo of the MRI-compatible core holder fabricated from Nitronic 60 stainless steel.

The core plug sample is held by heat shrink tubing and an Aflas polymer sleeve to make a connection to

the inlet and outlet fluid distribution plugs. The encapsulated sample is positioned inside the RF probe.

The entire assembly goes into the metal vessel. The vessel is sealed at the ends by o-rings on the plugs.

Radial pressure is exerted on the sample through the containment sleeve by pressurizing the internal

space. The Aflas sleeve and heat shrink isolate the sample from the pressurizing fluid. The internal space

is pressurized by introducing a fluid nonobservable to MR (e.g., fluorinated oil or gas) through external

connection by means of a hydraulic pump or gas cylinder; in this work nitrogen gas was chosen to

pressurize the core holder. Axial flow through the sample is provided by fluid pumped through fluid

distribution plugs at each end of the sample.

Material Choices.

As discussed earlier, an optimal material for construction of the body of the vessel will offer exceptional

tensile strength, low electrical conductivity, and relatively high thermal conductivity. The material with

the best balance of these characteristics was chosen to be an alloy of stainless steel, Nitronic 60

stainless steel (McMaster Carr, OH). It is an austenitic steel making it nonmagnetic, while still offering

the benefits of high tensile strength (400 MPa), low electrical conductivity (1.39 106 S/m) and

relatively high thermal conductivity (16 W/mK), compared to other possible materials shown in Table 1

[18,19].

For construction of the RF probe and body inside the pressure vessel, materials are chosen to minimize

hydrogen background signal. Early designs of the core holder showed issues with background signalfrom the materials used, specifically the epoxy and material used for construction of the RF probe as

shown in Fig. 2(b). For the first version, the epoxy employed was a marine epoxy (West Systems, MI) and

the RF probe body was made of G-10 fiberglass (McMaster Carr, OH) [7].

Figure 3(a) shows a 1D profile of the longitudinal (Z) axis of the first version core holder with a rock core

plug present. It is a core flooding experiment where water floods the rock from the left to right in the

image. It can be seen that the background signal is of the same order of magnitude as the sample. The

8/13/2019 Design Innovationdesign Innovation Papers

16/21

background signal originates with the marine epoxy and other polymers in the RF probe. The strength of

the background signal will be very temperature sensitive; the signal lifetime of polymers are generally

longer at elevated temperatures [26]. It is possible to record and average the background and then

subtract it from the experimental profiles. This methodology is relatively easy but not desirable; zero or

mean zero background signals would be ideal.

To further investigate the background signal a 2D image of the transverse plane (XY) of a water

saturated Berea rock plug was recorded and shown in Fig. 3(b). Note the bright spot at the center of the

rock core image. This is from water in the inlet/outlet tubing as seen in the side view image Fig. 2(a). At

the circular edge of the rock core, image signal is slightly enhanced due to background signal from the

Aflas sleeve housing the rock. There is also signal from epoxy in proximity to the eight rungs of the

birdcage RF probe, where the RF B1 field is strong [22].

For the next version of the core holder, modifications were made to minimize background signal. Glass-

filled PTFE (McMaster Carr, OH) instead of fiberglass G-10 was used for construction of the RF probe. As

part of this research we investigated a large number of potential epoxies to determine the most suitable

for construction of the RF probe. The MRI signal behavior of the epoxies was investigated to determine

the epoxy with the best combination of low signal magnitude and quickly decaying signal. The optimal

choice was found to be epoxy M31CL (McMaster Carr, OH). Figure 3(c) displays the results of these

efforts, showing the same transverse image of a saturated rock sample in the newer version of the

probe with reduced background signal. The ratio of signal from a water saturated rock sample to

background signal was measured for both versions of the core holder and increased from a ratio of

approximately 1:1 to 35:1 in the old versus new version. This shows how critical material selection is to

the process of designing and fabricating a MRI-compatible pressure vessel; especially in the RF probe.

Other Design Considerations.

The design of the core holder had many constraints on physical size. The outside diameter of the core

holder has a maximum limit which is the diameter of the gradient bore. The outside diameter had a

minimum limit as per the required thickness for a design pressure and sample size. The inside diameter

is limited by the size of the sample and the RF coil. The most common core holder sample sizes are 1.5

or 1 in. in diameter; 1 in. was chosen for this design. The RF probe requires a certain diametrical

clearance between the sample and vessel wall. For the core holder shown, the probe assembly was

roughly 1/2 in. thick. The length is constrained by the rock sample sizes as well as the fluid distribution

plugs. Core plugs are typically several inches in length, accommodation from 0 to 3 in. lengths were

considered in this design.

Different designs were considered and used for the sealing of the core holder on each end. Each end has

several longitudinal connections with lines/wires attached so the sealing mechanism cannot be one that

rotates or else the lines and connections will become tangled. Many commercial pressure vessels use a

sealing mechanism as outlined in Fig. 4(a). The size limitations of fitting inside the magnet bore prevents

the utilization of a large diameter bolt circle with a face-sealing o-ring, as it is not space efficient. The

sealing mechanism chosen for the core holder was an o-ring sealing similar to a piston, as shown in Fig.

8/13/2019 Design Innovationdesign Innovation Papers

17/21

4(b). This design proved to be most effective, having the benefits of no rotation of the end caps

required, as well as being more space efficient relative to conventional face-sealing pressure vessels.

As mentioned earlier, eddy currents and their induced magnetic fields were simulated for various

metallic pressure vessels. Simulation permits rapid survey of vessel parameters such as metal selection

and geometry, which would not be easily accomplished experimentally due to the associated materialand machining costs. Design of the vessel involved iteration between the simulations and geometry;

vessel geometry would be simulated and then modified according to the simulation results [7].

Imaging Results

Abstract |Introduction |MRI Background |MRI-Compatible Metallic Pressure Vessel Design |Design

and Constraints |Imaging Results |Conclusion |Acknowledgements |References

As a simple example of the nonmagnetic Nitronic 60 stainless steel core holder, water flooding in a

sandstone rock core was monitored with MRI. The two-dimensional centric scan Spiral SPRITE [14] was

employed for test imaging. As shown in Fig. 5(a), water was driven through a dry Berea sandstone fromright to left. The confining pressure and the temperature were set to 2.1 MPa (300 psi) and50 C,

respectively. A flow rate of 0.3 ml/min of water was used to flood the sample. During the flooding

experiment, 2D images were acquired at intervals of 43 s; the acquisition time per 2D image was 43 s

with four signal averages. After 40 min of imbibition the rock was fully saturated. Figure 5(b) shows 1D

profiles extracted from the centerline of the 2D image at intervals of 3.6 min.

Conclusion

Abstract |Introduction |MRI Background |MRI-Compatible Metallic Pressure Vessel Design |Design

and Constraints |Imaging Results |Conclusion |Acknowledgements |References

This paper discusses the development of high-pressure MRI-compatible metallic vessels. Compared to

conventional MRI core holders and pressure vessels, the proposed metallic pressure vessels permit

much more efficient temperature control, optimized sensitivity, and act as an EMI shield. Nitronic 60

stainless steel was found to be the most effective material for the vessel with a combination of high

tensile strength, relatively high thermal conductivity, and low electrical conductivity. Eddy currents are

dramatically reduced using stainless steel because of its relatively low electrical conductivity compared

to other metals. The sensitivity is greatly improved compared to conventional MRI compatible pressure

vessels, due to the probe being inside as opposed to outside the vessel.

Background signal from polymer materials inside the metal vessel are an issue, being on the same orderof magnitude as sample signal in earlier versions of the vessel. Simple averaging of the background

signal and subtraction was demonstrated as a short-term solution, but different material choices to

minimize background signal is a better strategy. Quantitative imaging of water flooding a rock core plug

in a stainless steel core holder under moderate pressure and temperature shows the potential of this

methodology. MRI compatible metallic pressure vessels enable macroscopic systems at high pressure

and variable temperature to be monitored by MRI.

8/13/2019 Design Innovationdesign Innovation Papers

18/21

Acknowledgements

Abstract |Introduction |MRI Background |MRI-Compatible Metallic Pressure Vessel Design |Design

and Constraints |Imaging Results |Conclusion |Acknowledgements |References

B.J.B. thanks the Canada Chairs program for a Research Chair in MRI of Materials (20092016), NSERC of

Canada for a Discovery grant, and ConocoPhillips. We thank many for their assistance and suggestions:

Dr. I. Mastikhin and Dr. B. Newling, as well as M. Olive and B. Titus. We thank Adam Melanson for his

investigation of water impermeable epoxies.

References

Abstract |Introduction |MRI Background |MRI-Compatible Metallic Pressure Vessel Design |Design

and Constraints |Imaging Results |Conclusion |Acknowledgements |References

1

Benedek, B., and Purcell, M. J., 1954, Nuclear Magnetic Resonance in Liquids Under High Pressure, J.

Chem. Phys., 22, pp. 20032012.[CrossRef]

2

Jonas, J., 1982, Nuclear Magnetic Resonance at High Pressure, Science, 216, pp. 11791184.[CrossRef]

[PubMed]

3

Grunwaldt, J. D., Wandeler, R., and Baiker, A., 2003, Supercritical Fluids in Catalysis: Opportunities of In

Situ Spectroscopic Studies and Monitoring Phase Behavior, Catal. Rev., 45, pp. 196.[CrossRef]

4

Angel, R. J., Ross, N. L., Seifert, F., and Fliervoet, T. F., 1996, Structural Characterization of Penta

Coordinate Silicon in a Calcium Silicate, Nature, 384, pp. 441444.[CrossRef]

5

Horvath, I. T., and Miuar, J. M., 1991, NMR Under High Gas Pressure, Chem. Rev., 91, pp. 13391351.

[CrossRef]

6

Lee, S. K., Mibe, K., Fei, Y. W., Cody, G. D., and Mysen, B. O., 2005, Structure of B2O3 Glass at High

Pressure: A 11B Solid-State NMR Study, Phys. Rev. Lett., 94, 165507-1, 4.[CrossRef]

7

8/13/2019 Design Innovationdesign Innovation Papers

19/21

Han, H., Ouellette, M., MacMillan, B., Goora, F., MacGregor, R., Green, D., and Balcom, B. J., High

Pressure Magnetic Resonance Imaging With Metallic Vessels, J. Magn. Reson., 213, pp. 9097.

[CrossRef][PubMed]

8

Han, H., Ouellette, M., MacGregor, R., Green, D., and Balcom, B. J., Non-Cartesian Sampled Centric Scan

SPRITE Imaging With Magnetic Field Gradient and B0 Field Measurements for MRI in the Vicinity of

Metal Structures, J. Magn. Reson., 206, pp. 97104.[CrossRef][PubMed]

9

Thurecht, K. J., Hill, D. J. T., and Whittaker, A. K., 2005, Equilibrium Swelling Measurements of Network

and Semi-Crystalline Polymers in Supercritical Carbon Dioxide Using High-Pressure NMR,

Macromolecules, 38, pp. 37313737.[CrossRef]

10

Morris, R. H., Bencsik, M., Nestle, N., Galvosas, P., Fairhurst, D., Vangala, A., Perrie, Y., and McHale, G.,

2008, Robust Spatially Resolved Pressure Measurements Using MRI With Novel Buoyant Advection-

Free Preparations of Stable Micro-Bubbles in Polysaccharide Gels, J. Magn. Reson., 193, pp. 159167.

[CrossRef][PubMed]

11

Baldwin, B. A., Stevens, J., Howard, J. J., Graue, A., Kvamme, B., Aspenes, E., Ersland, G., Huseb, J., and

Zornes, D. R., 2009, Using Magnetic Resonance Imaging to Monitor CH4 Hydrate Formation and

Spontaneous Conversion of CH4 Hydrate to CO2 Hydrate in Porous Media, Magn. Reson. Imaging, 27,

pp. 720726.[CrossRef][PubMed]

12

Li, L., Chen, Q., Marble, A. E., Romero-Zern, L., Newling, B., and Balcom, B. J., 2009, Flow Imaging of

Fluids in Porous Media by Magnetization Prepared Centric-Scan SPRITE, J. Magn. Reson., 197, pp. 18.

[CrossRef][PubMed]

13

Callaghan, P., 1993, Principles of Nuclear Magnetic Resonance Microscopy, Clarendon, New York.

14

Halse, M., Rioux, J., Romanzetti, S., Kaffanke, J., MacMillan, B., Mastikhin, I., Shah, N. J., Aubanel, E., and

Balcom, B. J., 2004, Centric Scan SPRITE Magnetic Resonance Imaging: Optimization of SNR, Resolution

and Relaxation Time Mapping, J. Magn. Reson., 169, pp. 102117.[CrossRef][PubMed]

15

8/13/2019 Design Innovationdesign Innovation Papers

20/21

Bernstein, M. A., King, K. F., and Zhou, X. J., 2004, Handbook of MRI Pulse Sequences, Elsevier Academic,

Burlington, VT.

16

Blumich, B., 2000, NMR Imaging of Materials, Oxford University Press, Oxford.

17

Timoshenko, S., 1956, Strength of Materials, Part II, Princeton University Press, Princeton.

18

Davis, J. R., 2008, ASM Handbook Committee, Metals Handbook, ASM International, Materials Park, OH.

19

Armco, 1990, Nitronic 60 Stainless Steel Bar and Wire UNS-S21800 Product Data Bulletin, Nitronic 60

stainless steel data sheet.

20

Grber, H., Erk, S., and Grigull, U., 1961, Fundamentals of Heat Transfer, McGraw-Hill, New York.

21

Hoult, D. I., 2000, The Principle of Reciprocity in Signal Strength CalculationsA Mathematical Guide,

Concepts. Magn. Reson., 12, pp. 173187.[CrossRef]

22

Mispelter, J., Lupu, M., and Briquet, A., 2006, NMR Probeheads for Biophysical and Biomedical

Experiments Theoretical Principles and Practical Guidelines, Imperial College Press, London.

23

Stoll, R. L., 1974, The Analysis of Eddy Currents, Clarendon, Oxford.

24

Yee, K. S., 1966, Numerical Solution of Initial Boundary Value Problems Involving Maxwell's Equations in

Isotropic Media, IEEE Trans. Antennas Propagat., 14, pp. 302307.

25

Han, H., MacGregor, R., and Balcom, B. J., 2009, Pure Phase Encode Magnetic Field Gradient Monitor,

J. Magn. Reson., 201, pp. 212217.[CrossRef][PubMed]

26

8/13/2019 Design Innovationdesign Innovation Papers

21/21

MacMillan, B., Halse, M., Schneider, M., Fardy, L., Chui, Y. H., and Balcom, B. J., 2002, Magnetic

Resonance Imaging of Rigid Polymers at Elevated Temperatures With SPRITE, Appl. Magn. Reson., 22,

pp. 247256.[CrossRef]