MES celebrates 80-year anniversary
Mitsui Engineering and Shipbuilding Co., Ltd. (MES), one of
Japan's leading heavy industrial companies in production of ships, steel structures,
process plants, machinery and equipment, is celebrating its 80-year anniversary
as a metal fabricator.
Since its inauguration in 1917 as a shipbuilder, the company
has steadily diversified its line of business along with Japan's industrial
development.
After establishing its chemical plant division in 1938, MES
achieved sound growth both as a plant process contractor and a process
equipment fabricator.
MES has fabricated a variety of equipment for petroleum
refining, petrochemical, synthetic resin, synthetic fiber, gas fertilizer,
organic or inorganic chemical processing plants.
MES has been fabricating common pressure vessels and heat
exchangers as well as more mechanized equipment such as rotary dryers and
agitating reactors. The company has handled many types of structural materials,
from carbon steel to zirconium.
MES started fabricating with advanced material equipment in
1956, when it began producing titanium vessels. These days, many chemical
companies are interested in special metals as corrosion resistant or heat
resistant materials that will ease the maintenance of their plants.
MES fabricated this large, solid zirconium column, 40m in
length and 1.6m/ 2.1m in diameter
Although these metals are known for their superiority
against corrosion and heat, it is said that fabrication, especially welding
special metals, requires much caution and a clean atmosphere. To meet such
requirements, MES established a "Clean Welding Shop" exclusively for
the fabrication of high alloy and non-ferrous metal equipment. In this 2000 m2
facility (established in 1982), main seams of vessels and heat exchangers
are welded using automated or mechanized welding systems, providing high
efficiency and uniform quality. In addition to the mechanized welding systems,
MES employs many skilled welders who have mastered the techniques needed to
work with special metals.
The shop has fabricated large, solid zirconium columns (see
photo above) and heat exchangers containing many tubes. The largest column the
company has ever fabricated weighs 40 tons, and its largest heat exchanger
contains 5,000 tubes.
MES is capable of designing and fabricating any kind of
zirconium heat exchanger, including tube-sheet types of solid zirconium,
explosion-bonded clad zirconium, and loose-lined zirconium.
Another feature of the shop is a pickling facility. Pickling
by strong acid containing hydrofluoric acid for completed titanium and
zirconium equipment is often required to remove free iron, which contaminates
the metal surfaces. This pickling facility enables MES to reply to any of its
customers' pickling requirements and provides contamination-free equipment.
MES is recognized as the most experienced fabricator of
zirconium in Japan, and the quality of its work has earned it a fine reputation
from many customers around the world.
For more information, call MES/Sales
Dept. in Tokyo (Attn: Mr. S. Yoneda) Tel: 81-3-3544-3362, Fax: 81-3-3544-3596.
Wah Chang's representative in Japan,
Mitsui & Co., Zirconium Section, Tel: 81-3-3285-3356, Fax: 81-3-3285-9985
CPI Applications Development Team Formed
Wah Chang's corrosion resistant Zircadyne Zirconium® equipment
has been very successful in a variety of chemical processing environments,
including acetic and nitric acid service. But there are producers of other
media, such as sulfuric acid where zirconium is only occasionally used, who
might benefit greatly from the metal's attractive properties.
To that end, Wah Chang is pleased to announce that it has
formed a new CPI Applications Development Team. The group, headed by Doug
Brenizer, will seek new applications for Zircadyne Zirconium® in
hydrochloric, phosphoric, sulfuric and other acids and will continue to develop
more mature niches, such as cetic and nitric acid.
Brenizer's team includes Wah Chang Process Industries expert
Rob Henson, whose background includes 12 years in the company's corrosion
laboratory. Wah Chang metallurgist Jack Tosdale, well known in ASTM circles,
joins the team as Manager of Corrosion Services. Rounding out the group is
Larry Duke, formerly a Sr. Reliability Engineer with Millennium Petrochemicals.
Duke, who recently joined Wah Chang, is a Mechanical Engineer and a member of
MTI.
Jack Tosdale
Larry Duke
Doug Brenizer
Rob Henson
Zirconium/Organics Conference Recap
Wah Chang recently hosted the first International Conference
on the use of Zirconium in Organic Acid Environments. Over 100 attendees, 23
speakers, and 17 exhibitors representing companies from 13 countries participated
in the three-day meeting at the Salishan Resort in Gleneden Beach, Oregon.
Kicking off the conference, Keynote Speaker B. J. Sanders of
Sanders USA paid tribute to the late Terry Webster, corrosion resistant metals
pioneer, Wah Chang metallurgist, and friend to many in the Chemical Processing
Industries.
Following that very powerful lead-in were many excellent
papers covering a variety of topics, including 'tissues Around the Use of
Zirconium in MMA Production" by Keith Briegel and Bart Frechem of Rohm
& Haas, "Issues Around the Use of Zirconium in Alcohols
Production" by Brian Fitzgerald of Exxon, and
"Performance of Zirconium in a Peracetic Acid
Distillation Unit" by Timo Korvela of Kemira, Finland.
Ralph Nauman, President of Wah Chang, closed the conference
with a speech that highlighted quality, delivery, and other improvements that
the company is making to its CPI zirconium products. Nauman emphasized that Wah
Chang would continue to work on shortening delivery times, describing in detail
the new equipment and off-site resources that will make it possible.
Following the conference, participants had the opportunity
to see some of the facility improvements first hand during an optional plant
tour and welding seminar held at the company's Albany, Oregon site.
Wah Chang is already making plans for a second meeting in
1999. To be put on a mailing list for conference updates, fax the company at
541-924-6892.
Forming zirconium heads and parts
By Jack Tosdale, Wah Chang and Richard Jenkins, formerly of
Phoenix
Introduction
The properties of the starting plate destined for forming
operations play a key role in successful fabrication of a part. Due to the
nature of the deformation systems in zirconium (one predominant slip and three
predominant twin systems), wrought product is highly anisotropic in various
properties affecting the forming. Yield strength, tensile strength, ductility,
and impact strength vary significantly with direction in the material. These
variables and their anisotropy must be considered in the design of the part and
the fabrication process to ensure an acceptable end product.
The plate rolling operation imparts a high compressive
stress normal to the plate being rolled. This stress results in a large tensile
strain in the rolling direction and a small tensile strain in the direction
transverse to the rolling direction. These compressive and tensile strains
impart a texture in the microstructure with the basal poles perpendicular to
the rolling direction and the basal plane tilted about 35° from the rolling
plane towards the transverse direction. As rolling progresses, this preferred
orientation becomes more intense1.
Initially, deformation produces twins, which reorient the
grains so that deformation by slip occurs, slip being the primary mechanism for
the major deformation. This texture happens because of zirconium's hexagonal
close-packed crystal structure and the deformation paths it provides. As a
result, there are significant differences in properties between the rolling,
transverse, and normal directions. Particularly affected are the tensile
properties, ductility, and impact toughness. These anisotropic properties can
result in serious problems during head forming, such as cracks and misshapen
heads, unless preventive measures are taken during the forming process.
This article reviews deformation and textures in plates and
the resultant effect of this texture on tensile properties, ductility and
impact resistance of zirconium. It also provides information on techniques to
mitigate the effects of anisotropy during the forming process.
Properties of Zirconium
Properties that promote the formability of zirconium
include:
1. Relatively low tensile strength and
good overall ductility
2. No ductile-to-brittle transformation at working temperatures.
3. A temperature effect that improves fracture sensitivity and impact toughness
and lowers working forces and stresses.
Factors that degrade the formability and hinder the head
forming process include:
1. Anisotropy of the crystal structure
and the resultant properties.
2. Low impact strength, which leads to brittle-type fractures and notch
sensitivity.
3. Reactivity with the interstitials (O, N, and H), causing hardening and
embrittlement.
4. Springback caused by the low modulus of elasticity.
5. Internal defects, such as inclusions, unrecrystallized grains, and porosity.
6. Surface defects, such as cracks, notches, scratches, and work hardened
areas.
When these properties are factored into the forming process,
chances of success are vastly improved. Probably the most significant factor
relating to acceptable formability is zirconium's low impact resistance and
high notch sensitivity. Impact resistance is a measure of the material's
ability to absorb high impact energy without fracturing. The condition complicating
this impact resistance has to do with the anisotropy of this property in
zirconium. Impact resistance varies from less than 50 ft-lbs to over 100
ft-lbs, depending on both sample and notch orientation. The following sections
describe first the texture imparted during rolling deformation, then the
anisotropy of tensile properties, ductility, and impact resistance.
Deformation and Textile
Zirconium develops a texture when rolled into plate and
sheet because of the deformation mechanisms available in a hexagonal
close-packed structure. Recrystallization of the structure will not
redistribute these textured grains into a random pattern, so the texture
remains. Deformation in zirconium is by both slip and twinning.
Slip occurs where specific planes of atoms oriented in the
slip direction, slide over each other. This mechanism does not change the
overall grain orientations, so it does not impart any texture. On the other
hand, twinning occurs primarily on three planes, all in the basal pole
direction. Crystals are rotated in this process and a texture is developed.
Because the twinning occurs in grains that have the basal
poles pointing within about 30° of the rolling direction, the resulting twinned
crystal bands are in a direction that is perpendicular to the rolling direction
and at a transverse angle of-35° from the normal to the rolling direction.
Temperature during deformation doesn't significantly affect the texture
formation, and the pole figures will be quite similar for all temperatures in
the alpha phase of zirconium (<860°C).
Effect on Tensile Properties and Ductility
Figure 1 presents a plot of the tensile properties as a
function of temperature. As expected, the strength decreases with temperature
while the elongation increases. As Figure 2 shows, the tensiles vary with
orientation; the transverse yield strength is higher than the longitudinal
(rolling) at all temperatures, while the opposite effect occurs for the
ultimate.
Table 1 also shows this anisotropy of tensile properties in
the two directions, longitudinal (rolling direction) and transverse, at room
temperature, based on Wah Chang production data.
These relations exist because of the texture effect. Higher
strengths occur when the basal poles are parallel to the stress direction. In
the textured structure after rolling, the basal poles are more aligned in the
transverse direction, so yield strength is higher there; however, ductility is
lower in this direction.
Effect on Impact Strength
The ability of a material to absorb impact energy and resist
crack propagation is a measure of its impact strength or toughness. The Charpy
test using a notched specimen is a common method to measure this property. For
zirconium, impact energies range from near zero to about 120 ft-lbs, depending
on chemical composition, fabrication history and specimen orientation relative
to the rolling direction (see Figure 3). The figure shows some difference in
impact resistance between the longitudinal and transverse directions. However,
there is a very large difference, as much as 30 ft-lbs, between the vertical
direction (across the thickness) and the horizontal (in the rolling plane).
In forming of plates into heads, cones or cylinders, there
are high tensile stresses in the rolling plane, where the impact resistance is
much less. As a result, if other measures are not taken, cracks will more
easily initiate on the outside rolling surface and especially on notches in the
rolling plane.
Fig.1. Tensile properties change with temperature
Fig.2. Tensiles vary with sample orientation
Fig.3. Impact strength as a function of temperature
Based on studies of fractured surfaces, it has been found
that zirconium fractures in a ductile rather than brittle manner; however,
because zirconium has such low resistance to impact, the fracture has the
appearance of a brittle fracture; so, it would be beneficial to include a
measure in the forming process to improve the impact strength and resistance to
fracture. The most important measure is temperature, as shown in Figure 3. This
plot shows that by increasing the forming temperature to 200°C (400°F), the
impact strength in the rolling plane can be increased to over 80 ft-lbs from
about 45 ft-lbs at room temperature.
Two factors that are very important in the response of
zirconium to an impact test are the notch characteristics and the speed of the
test or impact velocity. The brittle behavior as described, above is a result
of the low impact strength coupled with the notch and speed of impact. The
notch imposes a stress point and gradient that lowers the impact resistance
significantly. Lustman and Kerze2
state that an un-notched tensile specimen exhibited a reduction in area of 30%
while the notched one showed a 15% RA. In impact, the un-notched specimen
absorbed over 100 ft-lbs while the notched one absorbed only 15 ft lbs. This
shows that the notch is a very important contributor to the 'brittle' behavior
of zirconium.
Wah Chang recently performed a study where it put scratches
in the surface of samples, and the samples were then either pickled or heated
to oxidize the surface. Heating at 1050°F appears to help the room temperature
bendability; however, with the variability in the results, it is difficult to
draw firm conclusions. If anything, it appears that the oxidation treatments
may reduce the sensitivity of the plate to notches.
Part 2: Practical Aspects of Forming Zirconium Parts
The forming of zirconium heads and parts has always been
considered more of an art form than any approximation of an exact science. This
opinion is justified because of the small number of zirconium heads and formed
parts that are manufactured each year, spread out over several fabricators and
the lack of a large data base on which to develop controlling parameters
favorable to zirconium's metallurgical characteristics. This typically results
in a larger percentage of failures compared to stainless or even titanium heads
and parts.
This problem or concern can be reduced if the zirconium
producers and fabricators pool their data and talents to develop procedures for
forming zirconium that has improved success rates over our currently accepted
industry procedures. Presently, each fabricator has its own set of standards
for forming the zirconium parts or heads. If the heads are manufactured by a
head-forming vendor, then their own interpretations of what is needed are added
to this set of variables.
The remaining sections of this article look at the factors
that influence the forming characteristics of zirconium and determine what
actions producers and fabricators can take to enhance actual forming
operations.
Notch Sensitivity
Zirconium is an extremely notch sensitive metal. The
existence of scratches, grinder marks, dings, scrapes or other sources of
sharp-bottomed marks can have serious consequences if allowed to be stressed by
operations like bending, expanding or forming.
Zirconium plates that will be used for heads should be
thoroughly inspected for scratches and other defects. If any such defects are
found, they should be removed by grinding and polishing to remove the
sharp-bottomed area of the defect. If coarse media is used for the initial
grinding, further polishing to remove the scratches created by the coarse media
is needed. Pickling after the grinding is the preferred method but is not
typically available to most fabricators due to the environmental issues
involved. Consequently, polishing is the method normally available and should
be used to produce a smooth surface finish. The smoother finish results in
scratches that are not as deep and consequently have a reduced likelihood of
cracking during forming operations.
Typically, one side of the zirconium plate (the down side)
will have a better surface condition due to the nature of the hot rolling
process. Plates should be thoroughly examined on both sides. After
conditioning, the worst side should be used as the inside of the formed part.
This will place the worse surface problems into compression rather than tension
and should reduce tendencies for any scratches to open up to cracks during the
forming operation.
Impact Resistance
Zirconium has much lower impact resistance than typical
steel or stainless steels under almost all conditions. This places some
restrictions on the processes normally used by fabricators to manufacture
zirconium heads and parts.
For operations that produce an impact type of force on
zirconium, the speed of equipment operations typically should be reduced by
30-50% from carbon steel parameters as a starting point. Smaller reductions in
several steps should be used to attain the overall reductions needed to form
the parts or heads.
One way to improve impact resistance is to increase the
temperature of the metal during the forming operation. A working temperature of
400-600°F is typically used, but this is not to be considered an absolute upper
limit. For severe types of forming operations, temperatures can be up to 1100°F
to improve the formability and impact resistance of zirconium.
Anisotropic Properties
Zirconium is an anisotropic metal like titanium, which means
that properties vary according to the direction of the material. Tensile
properties, ductility and even thermal expansion can be markedly different
depending upon material direction.
This property means that particular care should be used to
maintain orientation of material direction for material that will be used to
form heads and other parts. This is very important when formed parts must be
produced from plates that are welded together to make the necessary starting
blank. The ideal method is to weld two equal pieces from the same plate with
the same material direction. This technique effectively eliminates problems
with anisotropic behavior. However, as today's equipment becomes larger and
larger, this is not always feasible from an economic or schedule requirement.
Consequently, careful planning should be used to lay out the zirconium starting
blanks to match material directions with each piece used in making the blank.
This can be a very important consideration in forming as a
mismatched blank can cause problems by deforming at differing rates within the
blank. This can result in asymmetric parts when that was not the intended
result. Flanges can be shorter than expected or 2:1 heads may be deformed in
some portion of the head.
Bending & Minimum Bend Diameter
Zirconium is a reasonably formable metal that can be readily
bent at a 5T bend radius with minimal problems. It is those minimal problems
that can ruin a fabricator's day and seriously affect a production schedule.
Zirconium can be readily bent at a 5T bend radius, and Wah
Chang typically tests all of its plate through 1/2 in. to a 5T bend radius.
Plates that are thicker than 1/2 in. are machined back on one side to 1/2 in.
and tested to a 5T bend radius. Why, then, do fabricators sometimes have
difficulty beading plates at radius greater than 5T? There are several common factors
affecting the bendability of zirconium.
First, temperature can greatly affect the minimum bend
radius. Do not attempt to bend zirconium plate to a minimal bend radius on a
cold day in January in an unheated shop with equipment that has been sitting
idle all weekend. Increasing the metal temperature greatly improves the
formability of the zirconium plate. Heating the plate and dies to 400-600°F
will improve the formability dramatically; however, it is not necessary to go
to this level of heating to improve the chances of forming parts. One can do
many simple things to increase the temperature without a great deal of effort.
Putting a heater or heat lamp on the zirconium plate and letting it heat up the
plates overnight will greatly improve the bending. The same thing can be done
with the dies. Increasing the temperature to 150-250°F can improve bending and
avoid problems with equipment that was not designed to operate at 400°F. You
will need special handling but not to the extent needed for a 400-600 degree
bending operation.
Second, verify your equipment setup against the actual
material. For example, bending 1/8-in. plate to a 5T bend radius typically
would mean using a 5/8-in. bend radius; however, the actual thickness of
1/8-in. plate can vary up to .150 in. thickness. This setup would actually bend
the plate to 4.15T bend radius. Adjust the setup to the actual material
thickness, or adjust the material thickness to the equipment setup.
Third, make sure the die setup is smooth and free of nicks
and scratches. Due to zirconium's extreme notch sensitivity, it is just as easy
to initiate a crack from a nick in the die/mandrel as it is to start a crack
from a deep scratch.
Fourth, select the recommended type of equipment for the
operation. In forming pipe, Wah Chang recommends that press break type
operations be limited to a maximum thickness of 1/4". Roll forming is
recommended for thicknesses over 1/4". (This guideline was developed over
many years of practical experience and should not be discarded or ignored. You
do so at your own risk.) If one must use a press break such as making 8-in.
Sch. 40 pipe, blanks should be examined closely for potential surface problems,
such as pits or scratches, and correct them before starting. Stress relieving
the blanks at 1050°F for 2-4 hours to produce a black oxide film on the blank
will also assist in forming by removing potential stresses and allowing the
black oxide to act as a lubricant during the forming operation.
Fifth, for U-tube exchangers, the recommended minimum bend
radius is 2D. It is possible to bend tubes at a smaller diameter, but several
factors come into play that affect the minimum bend radius. The initial factor
that affects the bend radius is wall thickness. Typically, one would not try to
bend a 20 BWG tube to a 1.5D radius, but it is easily done for a 14 BWG tube.
Depending upon the OD and wall thickness combination, it is possible to bend
zirconium tubes to a 1.5 D minimum bend radius with a high degree of success.
One thing that appears to greatly assist this minimum bend
radius is the air oxidized surface condition that Wah Chang supplies on the
longer tubes. This zirconium oxide layer on the surface appears to allow
bending to smaller bend radius than tubes with brightly annealed surface finish.
The tubes do not appear to fold as readily at the edge of the bending die or to
exhibit as much orange peel surface on the outer diameter of the tube. Again,
heating the tube and dies improves the bending. Use of a mandrel or packing the
tube with sand improves the bending radius. Sand may be a better choice as it
does not scar the ID of the tubes to any extent.
Handling of Welds in the Forming Processes
Welds in zirconium present special problems in the forming
process. It is easy to weld two pieces of zirconium plate together and then
bend the test sample on a mandrel to 105 degrees. Doing the same thing with a
zirconium head can present a number of problems because the actual operations
are doing more than bending the zirconium. There are several things that are
normally done to reduce the incidence of problems in welded zirconium head
blank.
The normal practice is to grind the weld bead flush with the
parent plate thickness. This is an acceptable practice. The weld bead is
readily visible on formed heads and typically looks to be a slight indentation
in the finished head. The weld bead is normally 0.001-0.005 in. below the
parent plate thickness. This indicates that the weld bead is actually deforming
more than the parent metal. This is the result of the heat-affected zone (HAZ)
being stronger than the center of the weld bead. This is probably a result of
the large grain structure in the center of the weld being worked against the
deformed grain structure in the heat-affected zone. The larger grains are not
as strong and readily deform during the forming process.
This effect can be reduced by leaving the weld slightly
higher than the parent plates. A gradual taper grind is needed to prevent any
sharp transition from parent plate to weld bead. Sharp transitions can act as
stress risers and cause cracking. Phoenix recently tried about a 10 percent
height increase for the weld bead on each side (e.g., a 1/2-in. plate should
have the weld bead about 0.050 in. higher than the parent plates) on two recent
heads that it had formed and produced acceptable results.
Cracking that does occur typically starts in the weld. This
is probably the result of residual stresses that remain from the welding
process. These stresses can be eliminated by stress relieving the welded blank
prior to the forming operation. Two recently welded heads that Phoenix formed
were stress relieved as part of the initial heating operation for head forming.
One of these heads was identical to two zirconium heads that Phoenix formed in
1996. Of the two heads formed in 1996, one cracked in the weld and had to be
repaired. The two recent zirconium heads experienced no problems and seemed to
validate the stress relieving operation as an acceptable method of reducing
potential problems.
The length of a weld will also have some impact on potential
problems and should be minimized.
Technician adjusts tube ends of a heat exchanger with a Zr
loose clad liner
Use the maximum width plate to reduce this length. It is not
necessary to use two equal sized pieces to manufacture a welded head blank. For
example, if a head blank is 120 in. in diameter, use one piece 96 in. by 120
in. and one piece 24 in. by 120 in. to produce a blank. This reduces the length
of the weld and lessens the cost of welding while reducing the risk of cracking
in the weld area during actual forming operations.
Springback or Memory Effects
Another interesting characteristic is the springback effect
wherein zirconium has a pronounced tendency to resist initial forming.
Zirconium parts cannot be readily formed to final size. The parts typically
must be overformed and allowed to springback to the correct dimensions. As an
example, to achieve a 90 degree bend, it is typically necessary to actually
bend the part to approximately 100-105 degrees and allow the part to return to
the 90 degree angle. This effect sometimes is time delayed. Parts will move
over a period of time. This effect is more pronounced in Zr 705 than in Zr 702.
This springback means that more effort must be used by the
equipment manufacturer in achieving dimensionally correct parts. Each operation
must be watched closely and more time is consumed in actual manufacture than
would normally be expected. Cylinders are initially rolled to size, welded,
then re-rolled to achieve correct size.
The actual amount of overforming needed is dependent on the
initial configuration of the starting blank, such as plate, the amount of deformation
and the finished configuration, such as a shell cylinder. Minimal amounts of
deformation actually increase the observed springback effect. There is no easy
rate to determine exactly how much overforming is required. This is a situation
where the experience factor is needed. New engineers may be able to better
explain the mechanisms involved in this memory effect and develop a role of
thumb to assist the fabricators in the future. For today, however, all of the
fabricators must rely on experienced people to guide them through this problem
and keep costs under control.
For more information
For more information about forming of zirconium heads and
parts, contact Wah Chang's Jack Tosdale at 541-917-6777.
References
1. ORNL-2944, "Metallurgy of Zir-caloy-2 Part 1, The
Effects of Fabrication Variables on the Anisotropy of Mechanical
Properties", P.L. Rittenhouse and M.L. Picklesimer, 1960, page 28.
2. "The Metallurgy of Zirconium", Benjamin Lustman
and Frank Kerzr, Jr., editors, McGraw-Hill Book Co., Inc., 1955, page 252.
