Preventing Crevice Corrosion in New

and Existing Steel Structures

• Eric Shoyer – Elzly Technology

• Pete Ault – Elzly Technology
• Pete McDonagh – Triborough Bridge and Tunnel Authority
• Brian Prazenka – Triborough Bridge and Tunnel Authority
 Background
For new structures, the state-of-the-art approach is to apply zinc-rich
primer to mating surfaces prior to assembly and to use galvanized
fasteners during assembly. Unfortunately, there are cases where this is not
always possible.
For existing structures, obviously it is typically not realistic to disassemble,
prime and re-assemble.
During maintenance painting, stripe coats and caulking are often used to
provide added protection to crevice areas. Though effective, caulking can
be expensive and may not be necessary on all surfaces.
 Introduction
Twent different test combinations were evaluated. These included
different types of surface preparation, fastener material, caulking extent,
and coating sequence were evaluated on aged and new steel assemblies.

The test assemblies were exposed to a cyclic accelerated corrosion test

and evaluated for rust staining, blistering, and pitting within the crevice
area.
 Test Panel Design
• 6” by 12” by 1/8” steel
• Three 2.5” by 4”steel coupons
• Assembled to create crevices for
water penetration
• One 4-inch long 2.5-inch Angle
• 10 Black Oxide bolts
(highlighted in blue box)
• 10 Galvanized bolts
(highlighted in red box)
 Test Matrix
Four different surface preparation Panel Condition
Assembly and Stripe
methods were evaluated in this System Surface Coat Caulk

project, they are: 1 No

New - Primed
None

Assembly
2

After

Coated
3 Top

Stripe
• New Steel – Assembled then SP- 4 3/4

10 5
6 No
Full

New - Primed
None

Assembly
7
• New Steel – SP- 10 as individual

Before

SP-10

Coated
8 Top

Stripe
parts, zinc primed then assembled 9
10
3/4
Full

• Weathered Steel – SP-10 prior to

11 No
None
12
coating

Coated
13 Top

Stripe
14 3/4

Weathered
• Weathered Steel – SP- 11 prior to 15
16 No
Full
None
coating 17

SP-11

Coated
18 Top

Stripe
19 3/4
20 Full
 Preparation of “Aged” Steel
• 10 panels were assembled/ coated/ pre- • Pre-Weathering Panel surface preparation:
weathered SP-11 Power Tool
• Coated with 3 mils of Epoxy SP-10 Near- Cleaning to Bare
• ~150 hours ASTM B117 Corrosion White Metal Metal using a
Exposure Blast Needle Gun &
• 9 months of Outdoor Exposure Grinding Wheel
(Vineland, NJ)
 New Steel Panels
• Two fabrication sequences
representing new steel:
• One set was assembled, abrasive
blasted (SP-10), and coated with a
OZ/E/URE System
• One set was abrasive blasted (SP-
10), OZ primed, assembled, and
then finish coated (EP/URE)
• All bolts where scuff sanded
prior to intermediate coating
 Coating Application
Coating Application procedure:
• Perform designated surface preparation (e.g., SP-10 or SP-11 per Test
Matrix)
• SSPC SP-1, Solvent Cleaning (using Isopropyl alcohol)
• Zinc Primer (3-5 mils)
• Zinc Stripe (3-5 mils)
• Intermediate Coat (3-5 mils)
• Intermediate Stripe (3-5 mils)
• Caulk application
• Finish Stripe (3-5 mils)
• Finish Coat (3-5 mils)
 Caulking Scenarios
Three panels from each surface
preparation scenario had caulk
applied, to determine the
effectiveness of the caulk at
preventing crevice corrosion
from occurring.
The three scenarios were:
• Full Method Full Method 3/4 Method Top Method
Caulking
• ¾ Method Caulking
applied to all applied to all
Caulking only
applied to top
• Top Method crevices except bottom crevices
crevices
 Performance Testing
• Corrosion (GMW14872)
• 120 cycles – Underbody method C
• Inspections 20/40/80/120
cycles
• ASTM D-610 (rusting)
• ASTM D-714 (blistering)
• Crevice corrosion locations
• Rusting bolt count
• Note: panels are rotated
positions every 20 cycles
• At end of test, panels were
disassembled for pitting analysis
within crevices
 Quick Review
• 10 panels were assembled/ coated/ pre-weathered
• Coated with 3 mils of Epoxy
• 150 hours ASTM B117 Corrosion Exposure + 9 months of Outdoor Exposure (Vineland, NJ)
• After exposure panels were prepped one of two methods:
• SP-10
• SP-11
• 10 panels represent two fabrication sequences for new steel:
• 5 panels were assembled, abrasive blasted (SP-10), and coated with an OZ/EP/URE System
• 5 Panels were abrasive blasted (SP-10), OZ primed, assembled, and then finish coated (EP/URE).
Note: All bolts where scuff sanded prior to intermediate coating
• Three different Caulking methods were used post coating application for each preparation scenario
• All Panels were then exposed to 120 cycles of GMW 14872 testing

Results –New Steel
20 cycles 40 cycles 80 cycles 120 cycles
20 cycles 40 cycles 80 cycles 120 cycles

• Primed • Primed
After No Stripe Before No Stripe

Assembly
+
No Caulk
Assembly +
No Caulk

Stripe Stripe
+ +
No Caulk No Caulk

Stripe Stripe
+ +
Top Caulk Top Caulk

Stripe Stripe
+ +
¾ Caulk ¾ Caulk

Stripe Stripe
+ +
Full Caulk Full Caulk


Results –Weathered Steel
20 cycles 40 cycles 80 cycles 120 cycles 20 cycles 40 cycles 80 cycles 120 cycles

• SP-10 No Stripe
• SP-11 No Stripe
+ +
No Caulk No Caulk

Stripe Stripe
+ +
No Caulk No Caulk

Stripe Stripe
+ +
Top Caulk Top Caulk

Stripe
Stripe
+
+
¾ Caulk
¾ Caulk

Stripe Stripe
+ +
Full Caulk Full Caulk

 Results – Black Oxide Bolts
• Most bolt corrosion were Front of
20
Back of Front of
40
Back of Front of
80
Back of Front of
120
Back of
panels primed prior to Panel Condition
Black Bolts Black Bolts Black Bolts Black Bolts Black Bolts Black Bolts Black Bolts Black Bolts

assembly. No 60% 100% 10% 100% 100% 100% 100% 100%

After Assembly
None

New - Primed
10% 0% 30% 10% 30% 10% 40% 50%

Stripe Coated
• The black-oxide bolts Top
3/4
10%
40%
40%
20%
10%
50%
40%
20%
20%
50%
40%
20%
30%
50%
80%
50%
that did not receive a Full 20% 0% 20% 0% 30% 0% 40% 40%

zinc primer or zinc stripe No 100% 100% 100% 100% 100% 100% 100% 100%

Before Assembly
None

New - Primed
0% 0% 0% 0% 80% 0% 90% 100%
coat displayed corrosion

Stripe Coated
SP-10
Top 20% 0% 30% 30% 70% 30% 100% 100%

at the first inspection 3/4

Full
0%
10%
0%
0%
40%
20%
30%
10%
70%
70%
30%
10%
100%
90%
100%
100%
(cycle 20). No
None
70% 100% 70% 100% 70% 100% 70% 100%
0% 0% 0% 0% 0% 0% 0% 30%
• Less evident on abrasive
Stripe Coated
Top 10% 10% 10% 10% 20% 10% 20% 20%

blasted panels than the 3/4 0% 10% 10% 10% 10% 10% 20% 10%
Weathered

power tool cleaned

Full 20% 0% 20% 0% 20% 0% 20% 0%

No 40% 100% 80% 100% 90% 100% 100% 100%

panels. None
20% 0% 30% 0% 30% 0% 30% 70%
Stripe Coated
SP-11

Top 10% 40% 30% 40% 30% 40% 30% 90%

3/4 20% 20% 50% 40% 60% 40% 80% 100%
Full 0% 20% 20% 50% 20% 50% 20% 70%
 Results –Galvanized Bolts 20 40 80 120

• Corrosion on the was Front of Back of Front of Back of

Galvanized Galvanized Galvanized Galvanized
Front of
Galvanized
Back of
Galvanized
Front of Back of
Galvanized Galvanized

less on galvanized Panel Condition

No
Bolts
0%
Bolts
0%
Bolts
50%
Bolts
100%
Bolts
70%
Bolts
100%
Bolts
70%
Bolts
100%

After Assembly
New - Primed
None
bolts than the 0% 0% 0% 0% 10% 20% 10%

Stripe Coated
0%
Top 0% 0% 0% 10% 0% 30% 10% 30%

corroded black-oxide 3/4

Full
0%
0%
0%
0%
0%
0%
0%
0%
10%
20%
10%
0%
20%
30%
10%
10%

bolts. No 0% 0% 10% 100% 100% 100% 100%

Before Assembly
50%

New - Primed
None

Stripe Coated
0% 0% 0% 0% 0% 0% 20% 40%

SP-10
• There are clear Top
3/4
0%
0%
0%
0%
0%
0%
20%
20%
0%
20%
30%
50%
20%
20%
40%
50%

benefits to stripe No
Full 0%
0%
0%
0%
30%
20%
20%
80%
50%
40%
40%
100%
50%
40%
40%
100%

coating the galvanized

None

Stripe Coated
0% 0% 10% 0% 10% 0% 10% 30%
Top 0% 10% 0% 10% 0% 20% 0% 20%

bolts. 3/4 10% 0% 10% 10% 10% 10% 20% 10%

Weathered

Full 0% 0% 0% 0% 0% 10% 0% 20%

No 0% 0% 20% 70% 20% 90% 20% 90%
None
0% 0% 0% 30% 0% 30% 10%
Stripe Coated

20%
SP-11

Top 0% 0% 0% 10% 0% 10% 0% 10%

3/4 0% 0% 20% 0% 20% 0% 20% 10%
Full 0% 0% 0% 0% 0% 10% 0% 30%
 Results - Disassembly
• ¾ caulking applied to new steel that is No Stripe
+
Stripe
+
Stripe
+
Stripe
+
Stripe
+

primed after assembly is the best way No Caulk No Caulk Top Caulk ¾ Caulk Full Caulk

to prevent corrosion in these crevices.

• In one of the two crevice areas, the
New-Primed
After

full caulking appears to hold moisture

Assembly

within the crevice of the new steel

panels.
• Panels with neither a stripe coat nor New-Primed
Before

caulking experienced the most Assembly

crevice corrosion.

• For weathered steel under repair

conditions, the benefits of a full caulk Weathered –

system can be observed.

SP-10
Remediation

• The remaining caulking schemes

visually appear better than the

schemes without caulking. Weathered –

SP-11
Remediation

 Pit Depth Analysis
• In an attempt to better quantify the crevice corrosion, pit
depths were measured on two different surfaces.
• After disassembling the panels, all corrosion products
were removed from the crevice surface of the angle using
abrasive glass bead blasting.
• Ten measurements were made to find the highest pits for
each surface
 Pit Depth
• L - Bracket • Larger Plate
 Results – Pit Depth – L Bracket
New Steel
10

Avg/Max/Min Pit Depth - mils

9
8
Weathered Steel 7
6
25 5
Avg/Max/Min Pit Depth - mils

4
20 3
2
15 1
0
10 Nothing Stripe Only Stripe and Caulk Stripe and 3/4- Stripe and Full
Top Caulk Caulk
5
Prime Before Assembly Prime After Assembly
0
Nothing Stripe Only Stripe and Caulk Stripe and 3/4- Stripe and Full
Top Caulk Caulk

SP-10 (Abrasive Blast) SP-11 (Needle Gun and Angle Grinder)

 Results – Pit Depth – Middle Crevice
New Steel
45

Avg/Max/Min Pit Depth - mils

Weathered Steel 40
35
60
30
Avg/Max/Min Pit Depth - mils

50 25
40 20
15
30
10
20 5
0
10
Nothing Stripe Only Stripe and Caulk Stripe and 3/4- Stripe and Full
0 Top Caulk Caulk
Nothing Stripe Only Stripe and Caulk Stripe and 3/4- Stripe and Full
Top Caulk Caulk Prime After Assembly Prime Before Assembly
SP-10 (Abrasive Blast) SP-11 (Needle Gun and Angle Grinder)
 Conclusions
• Galvanized bolts perform better over time than black oxide bolts.
• If black oxide bolts are utilized, proper surface preparation along with additional stripe
coats will help prevent corrosion from occurring.
• Stripe coats and caulking of crevices directly exposed to water/ moisture
will help prevent crevice corrosion on new steel.
• When caulking newly applied steel consider leaving the bottom crevice uncaulked to
allow moisture to escape.
• When working with weathered steel, full stripe coats and caulking of all
crevices provided the best results in regards to reducing crevice
corrosion and pitting.
• As a best practice, mating steel surfaces should receive a primer coating
prior to assembly
 Acknowledgements
This material is based upon work supported by Triborough Bridge and Tunnel
Authority (New York, NY) and performed by Elzly Technology. Any opinions,
findings and conclusions or recommendations expressed in this material are
those of the author(s) and do not necessarily reflect the views of the Triborough
Bridge and Tunnel Authority.
 Thank you
Any Questions?
Email: eshoyer@elzly.com
 “Preventing Crevice Corrosion in New and Existing Steel Structures”

By Eric Shoyer

Presented at SSPC 2018

January 15 – 18 , 2018
New Orleans, LA

Notice: This paper was presented by the author(s) or assigned speakers at the SSPC 2018 conference as indicated above. SSPC:
The Society for Protective Coatings ("SSPC") has a worldwide, royalty-free, fully paid up, perpetual, and irrevocable limited
license (with the right to sublicense) to do any and all of the following: Publish this paper in the official proceedings for the
conference; Record the related presentation on film, tape, disk or other forms of media for sale; Publish the paper or
presentation in the Journal of Protective Coatings and Linings; SSPC reserves the right of first publication of the paper or
presentation; Distribute printed copies of your presentation on-site to meeting attendees.

In addition, SSPC shall have the right to sublicense to its third party designees some or all of the rights as set forth above, in the
sole and exclusive discretion and under the direction and control of SSPC. As such, distribution or sharing of this content in any
print, digital or verbal format is not permitted without the consent of SSPC.
 Preventing Crevice Corrosion in New and Existing Steel Structures

Eric Shoyer
Elzly Technology Corporation
Vineland, NJ, USA

Pete Ault
Elzly Technology Corporation
Vineland, NJ, USA

Peter McDonagh and Brian Prazenka

Triborough Bridge and Tunnel Authority
New York, NY, USA

Abstract:

Steel structures are often fabricated by bolting (or riveting) structural components together.
Bolted joints result in multiple crevices between the bolted members and fasteners. These
crevices are typically more susceptible to corrosion than flat surfaces because they tend to
retain water/contaminants, they are difficult to properly coat, and the crevice geometry tends
to support electrochemical phenomena that accelerate corrosion. This paper will evaluate the
effectiveness of various coating practices at mitigating corrosion around these joints.

Introduction

For new structures, the state-of-the-art approach is to apply zinc-rich primer to mating surfaces
prior to assembly and to use galvanized fasteners during assembly. Unfortunately, there are
cases where this is not always possible. Obviously it is typically not realistic to disassemble,
prime and re-assemble existing structures. During maintenance painting, stripe coats and
caulking are often used to provide added protection to crevice areas. Though effective,
caulking can be expensive and may not be necessary on all surfaces.

This paper presents the results of a laboratory study investigating the degree of crevice
corrosion protection provided to bolted joints using different coating schemes. Forty different
combinations of surface preparation, fastener coating, caulking extent, and coating sequence
were evaluated on aged and new steel assemblies. The test assemblies were exposed to a
cyclic accelerated corrosion test and evaluated for rust staining, blistering, and pitting within
the crevice area. The data presented will help quantify the benefits of alternative approaches
for corrosion mitigation in fabricated steel joints.
 Test Approach

Panel Design

Complex panels were assembled to replicate various geometries and include various materials
that are seen on steel structures. A steel panel 6”x12”x1/8” had three 4”x2.5”x1/8” steel plates
fastened to them with ½” nuts and bolts (10 black steel and 10 galvanized) and a L channel steel
plate 4”x 2” x 1/8”. The panels are made in such a way that they have both vertical and
horizontal water travel paths, and an imitation built up joint.

Front of Panel Side View Angle View

Figure 1. Test Panel Configuration

There were a total of twenty panels for this project. To compare effects of new steal vs. aged
steel, half of the test panels were weathered prior to surface preparation and coating
application. Four different surface preparation methods were evaluated in this project, they
are:
• New Steel – Assembled then SP-10
• New Steel – SP- 10 as individual parts, zinc primed then assembled
• Weathered Steel – SP-10 prior to coating
• Weathered Steel – SP- 11 prior to coating
Figure 2 shows the test matrix for the study. The existing steel, new steel, and caulking
methods are described in the following sections.
 Panel Condition
Assembly and Stripe
System Surface Coat Caulk
1 No

New - Primed
None

Assembly
2

After

Coated
3 Top

Stripe
4 3/4
5 Full
6 No

New - Primed
None

Assembly
7

Before

SP-10

Coated
8 Top

Stripe
9 3/4
10 Full
11 No
None
12

Coated
13 Top

Stripe
14 3/4
Weathered

15 Full
16 No
None
17
SP-11

Coated
18 Top
Stripe

19 3/4
20 Full
Figure 2. Test Matrix

Weathered Steel

Prior to exposure, panels were coated with a light layer of epoxy coating (1-2 mils) over bare
steel with no surface profile. Panels were exposed for 150 hours of salt spray exposure1 and 9
months of outdoor exposure in an industrial setting. Figure 3 shows panels after weathering
but prior to surface preparation procedures.

Front Back

1
ASTM B117, Standard Practice for Operating Salt Spray
Figure 3. Panels after Weathering
After weathering, these panels had surface preparation to remove all previous coating and
corrosion by performing either one of two methods. The first was SSPC SP-10/NACE No. 2,
Near-White Blast Cleaning using 30 grit aluminum oxide to achieve a nominal 1-2 mil profile.
The second preparation method for weathered steel was SSPC SP-11, Bare Metal Power Tool
Cleaning using a combination of needle gun and angle grinder to achieve a nominal 1 mil
profile. Figure 4 shows representative weathered steel panels prepared for each scenario.

SP-10- “Near White SP-11- “Power Tool

Metal Blast” Cleaning to Bare Metal”
Figure 4. Surface Preparation after Weathering

New Steel

The second set of the test panels represented new steel assemblies. Half of these panels were
assembled prior to abrasive blasting and priming with a zinc-rich primer. The other half was
abrasive blasted as individual components, primed using the same zinc-rich primer and then
assembled. Figure 5 shows representative new steel panels after assembly and preparation.
All abrasive blasting was SSPC SP-10/NACE No. 2, Near-White Blast Cleaning using 30 grit
aluminum oxide to achieve a nominal 1-2 mil profile. The fasteners which on the panels that
were primed prior to assembly were hand sanded prior to additional painting.
 Assembled Prior to Blast Post SP-10 Blast SP-10 & Primed then assembled
Figure 5. New Steel Panels

Coating Application

The coating system used for this project was a standard three-coat system currently used on
structural bridges today. The coating application sequence was as follows for all panels. One
panel from each scenario did not receive stripe coats of each coating and two of each surface
preparation scenario did not receive caulking. Coating Application procedure:
• Perform designated surface preparation (e.g., SP-10 or SP-11)
• SSPC SP-1, Solvent Cleaning (using Isopropyl alcohol)
• Zinc Primer (3-5 mils)
• Zinc Stripe (3-5 mils)
• Intermediate Coat (3-5 mils)
• Intermediate Stripe (3-5 mils)
• Caulk application
• Finish Stripe (3-5 mils)
• Finish Coat (3-5 mils)

Caulking Methods

Three panels from each surface preparation scenario had caulked applied, to determine the
effectiveness of the caulk at preventing crevice corrosion from occurring. Of the three panels,
the caulking was applied in a different way to determine how much was needed to prevent
crevice corrosion. Figure 6 illustrates these different caulking scenarios.

Full Method ¾ Method Top Method

Figure 6. Caulking Scenarios – Side View
Corrosion Testing

Following a full 7-day cure after coating application, panels were exposed to 120 cycles of GMW
14872 testing. This testing is described below in figure 8.

Figure 7. GMW14872 One Cycle

Inspections were carried out at 20, 40, 80, and 120 cycles. Each panel was inspected for rust
through (ASTM D610), blistering (ASTM D714), crevice corrosion, and percentage of bolts with
corrosion. Following 120 cycles of corrosion testing, panels were disassembled to determine
the extent of pitting in each crevice.
 Results

Results and photos from the inspection intervals from each type of surface preparation are
shown below. Figure 8 shows the panels that were primed after assembly. Overall corrosion of
the panels at each cycle was very minimal with the exception of rust bleed from the crevices in
three of the panels. Crevice corrosion can be seen along the sides and top of the L-channel
early on throughout the exposure testing for the first three panels in this set.

20 cycles 40 cycles 80 cycles 120 cycles

No Stripe
+
No Caulk

Stripe
+
No Caulk

Stripe
+
Top Caulk

Stripe
+
¾ Caulk

Stripe
+
Full Caulk

Figure 8. New Steel–Primed after assembly
Figure 9 shows the panels that were primed before assembly. The black steel bolts on these
panels showed the most corrosion of the conditions evaluated. The panel that received no
stripe coating showed the greatest bolt corrosion. These bolts were only protected by the
epoxy intermediate coat as they were neither primed or stripe coated with the zinc-rich primer.
The remaining systems received a stripe coat of zinc-rich primer on the black steel bolts yet still
experienced corrosion around the black oxide bolts much earlier than any of the other types of
surface preparation systems evaluated. Other than the black oxide bolts, crevice corrosion was
less evident on the panels that were primed before assembly.
 20 cycles 40 cycles 80 cycles 120 cycles

No Stripe
+
No Caulk

Stripe
+
No Caulk

Stripe
+
Top Caulk

Stripe
+
¾ Caulk

Stripe
+
Full Caulk

Figure 9. New Steel - Primed after assembly

Not surprisingly, the abrasive blasted repair system had less bolt and crevice corrosion than the
power tool cleaned system. Figure 10 shows the abrasive blasted panels and Figure 11 shows
the power tool cleaned panels. Running rust can be seen from the crevices on the first few
panels where stripe and caulking do not cover this areas. With both types of surface
preparation, the best overall performing system was that of the full caulking; running rust was
not evident after corrosion testing. Varying degrees of running rust can be seen from the
crevices on the remaining panels.
 20 cycles 40 cycles 80 cycles 120 cycles

No Stripe
+
No Caulk

Stripe
+
No Caulk

Stripe
+
Top Caulk

Stripe
+
¾ Caulk

Stripe
+
Full Caulk

Figure 10. Weathered Steel – SP-10 Remediation
 20 cycles 40 cycles 80 cycles 120 cycles

No Stripe
+
No Caulk

Stripe
+
No Caulk

Stripe
+
Top Caulk

Stripe
+
¾ Caulk

Stripe
+
Full Caulk

Figure 11. SP-11 Weathered Steel – SP-11 Remediation

Crevice Corrosion

The extent of crevice corrosion was noted at each inspection during exposure testing. The top,
sides, and bottom of the three-coupon assembly along with the top, bottom, and sides of the
angle were inspected for evidence of running rust or rust through of the coating at these areas.
Figure 12 shows the extent of crevice corrosion after 120 cycles of exposure testing (red
indicates that corrosion was present).
 Running Rust at Crevices
Coupon Assembly Angle
System Panel Condition Top Sides Bottom Top Sides Bottom
1 No cv X X cv

New - Primed New - Primed

None

Assembly
2 cv X cv

After

Stripe
Coated
3 Top cv cv
4 3/4
5 Full
6 No cv cv cv cv X cv
None
Assembly

7 cv X X
Before

SP-10

Stripe
Coated
8 Top cv cv
9 3/4
10 Full cv
11 No X X cv X X X
None
12 Stripe X X X X cv
13 Coated Top cv cv X
14 3/4 cv cv
Weathered

15 Full cv
16 No X X X X X X
None
17 X X cv X X
SP-11

Stripe
Coated

18 Top cv cv cv
19 3/4 cv cv
20 Full
Figure 12. Crevice Corrosion inspection at 120 cycles of GMW14872 (red indicates that corrosion was present)

The only panels that did not display some form of crevice corrosion were those from the ¾ and
full-caulking scenarios. It was interesting to note that new steel panels that were primed after
assembly displayed less crevice corrosion than that of the primed before assembly panels,
though this may be influenced by the poorly coated black bolts.

Bolt Corrosion

To explore the effects of the coating techniques on galvanized and black-oxide bolts, the
percentage of bolts experiencing corrosion was documented at each inspection cycle. Figure 13
shows the percentage of black-oxide bolts displaying corrosion at each inspection. Figure 14
shows the same data for the galvanized bolts. Comparison of Figure 13 and Figure 14 clearly
shows the value of galvanized bolts in any scenario.

The panels displaying the most bolt corrosion were panels primed prior to assembly. The black-
oxide bolts that did not receive a zinc primer or zinc stripe coat displayed corrosion at the first
inspection (cycle 20). Epoxy and polyurethane is not sufficient to protect the black-oxide
fasteners for even a short period of time. If black-oxide bolts must be used, they should receive
full and stripe coats of zinc primer, epoxy intermediate and finish coat.

Corrosion of the weathered black-oxide bolts was less evident on abrasive blasted panels than
the power tool cleaned panels. This is not surprising, as existing rust in crevices was more
thoroughly removed using abrasive blasting than power tools.

When comparing Figure 13 and Figure 14, it is worth noting that the extent of corrosion on the
individual fastener was less on the corroding galvanized bolts than the corroded black-oxide
bolts. Even though galvanized bolts had significantly less corrosion than the black-oxide bolts,
there are clear benefits to stripe coating the galvanized bolts.
 20 40 80 120
Front of Back of Front of Back of Front of Back of Front of Back of
Black Bolts Black Bolts Black Bolts Black Bolts Black Bolts Black Bolts Black Bolts Black Bolts
Panel Condition
No 60% 100% 10% 100% 100% 100% 100% 100%
After Assembly

None
New - Primed

10% 0% 30% 10% 30% 10% 40% 50%

Stripe Coated Top 10% 40% 10% 40% 20% 40% 30% 80%
3/4 40% 20% 50% 20% 50% 20% 50% 50%
Full 20% 0% 20% 0% 30% 0% 40% 40%
No 100% 100% 100% 100% 100% 100% 100% 100%
Before Assembly

None
New - Primed

0% 0% 0% 0% 80% 0% 90% 100%

Stripe Coated
SP-10

Top 20% 0% 30% 30% 70% 30% 100% 100%

3/4 0% 0% 40% 30% 70% 30% 100% 100%
Full 10% 0% 20% 10% 70% 10% 90% 100%
No 70% 100% 70% 100% 70% 100% 70% 100%
None
0% 0% 0% 0% 0% 0% 0% 30%
Stripe Coated

Top 10% 10% 10% 10% 20% 10% 20% 20%

3/4 0% 10% 10% 10% 10% 10% 20% 10%
Weathered

Full 20% 0% 20% 0% 20% 0% 20% 0%

No 40% 100% 80% 100% 90% 100% 100% 100%

None
20% 0% 30% 0% 30% 0% 30% 70%
Stripe Coated
SP-11

Top 10% 40% 30% 40% 30% 40% 30% 90%

3/4 20% 20% 50% 40% 60% 40% 80% 100%
Full 0% 20% 20% 50% 20% 50% 20% 70%

Figure 13. Corrosion of Black Oxide bolts on Panels

20 40 80 120
Front of Back of Front of Back of Front of Back of Front of Back of
Galvanized Galvanized Galvanized Galvanized Galvanized Galvanized Galvanized Galvanized
Panel Condition Bolts Bolts Bolts Bolts Bolts Bolts Bolts Bolts
No 0% 0% 50% 100% 70% 100% 70% 100%
After Assembly
New - Primed

None
0% 0% 0% 0% 10% 20% 10%
Stripe Coated

0%
Top 0% 0% 0% 10% 0% 30% 10% 30%
3/4 0% 0% 0% 0% 10% 10% 20% 10%
Full 0% 0% 0% 0% 20% 0% 30% 10%
No 0% 0% 10% 100% 100% 100% 100%
Before Assembly

50%
New - Primed

None
Stripe Coated

0% 0% 0% 0% 0% 0% 20% 40%
SP-10

Top 0% 0% 0% 20% 0% 30% 20% 40%

3/4 0% 0% 0% 20% 20% 50% 20% 50%
Full 0% 0% 30% 20% 50% 40% 50% 40%
No 0% 0% 20% 80% 40% 100% 40% 100%
None
Stripe Coated

0% 0% 10% 0% 10% 0% 10% 30%

Top 0% 10% 0% 10% 0% 20% 0% 20%
3/4 10% 0% 10% 10% 10% 10% 20% 10%
Weathered

Full 0% 0% 0% 0% 0% 10% 0% 20%

No 0% 0% 20% 70% 20% 90% 20% 90%
None
0% 0% 0% 30% 0% 30% 10%
Stripe Coated

20%
SP-11

Top 0% 0% 0% 10% 0% 10% 0% 10%

3/4 0% 0% 20% 0% 20% 0% 20% 10%
Full 0% 0% 0% 0% 0% 10% 0% 30%
Figure 14. Percentage of galvanized bolts displaying corrosion per panel
Disassembly

After accelerated corrosion testing, panels were disassembled for analysis of corrosion in the
crevices. Figure 15 shows each panel dissembled.

No Stripe Stripe Stripe Stripe Stripe

+ + + + +
No Caulk No Caulk Top Caulk ¾ Caulk Full Caulk

New-Primed
After
Assembly

New-Primed
Before
Assembly

Weathered –
SP-10
Remediation

Weathered –
SP-11
Remediation

Figure 15. Panels disassembled after application.
Figure 16 shows that ¾ caulking applied to new steel that is primed after assembly is the best
way to prevent corrosion in these crevices. In one of the two crevice areas, the full caulking
appears to hold moisture within the crevice of the new steel panels. As expected, the panels
with neither a stripe coat nor caulking experienced the most crevice corrosion. The remaining
three scenarios (stripe coat, top caulk and full caulk) had visually similar crevice corrosion on
the new steel assemblies.

For weathered steel under repair conditions, the benefits of a full caulk system can be
observed. The remaining caulking schemes visually appear better than the schemes without
caulking.

Pit Depth

In an attempt to better quantify the crevice corrosion, pit depths were measured on the surface
of the angles that were mounted onto the larger panel. After disassembling the panels, all
corrosion products were removed from the crevice surface of the angle using abrasive glass
bead blasting. Figure 16 shows the crevice surface of an angle before and after glass bead
blasting.

L – Channel after disassembly L - Channel after blast

Figure 16. L channel before and glass bead blasting

Once the surface was cleaned, the ten deepest pits were measured using a Pit & Crack Depth
Gauge (Albuquerque Industrial, Inc). The instrument has a measuring range of 0 to 500 mils
and a resolution of 0.5 mils. The minimum, average, and maximum measurement are used for
the analysis presented in this paper. Figure 17 shows the data for the panels representing
painting of new steel. Figure 18 shows the data for the panels representing maintenance
painting of weathered steel.

Comparing the two graphs, it is clear that deeper pits were measured in the panels
representing maintenance painting of weathered steel. However, a certain portion of that
pitting occurred when the test panels were being weathered (i.e., pre-rusted). For any given
panel, it is not possible to know how much of the pitting occurred prior to the maintenance
painting. Analysis of the remaining crevice surfaces may provide sufficient statistical data to get
additional insight, however this paper will focus on the two data sets individually.
For the new steel painting scenarios (Figure 17), the pit depth data corroborates the visual
observations. Pitting is clearly reduced by striping and caulking. The caulking scenarios which
left at least one edge un-caulked had the least pitting. Priming prior to assembly was also
beneficial in these scenarios. Interestingly, priming prior to assembly led to greater pitting than
priming after assembly for the scenarios which did not have a stripe coat or were fully caulked.
Standard industry practice of priming surfaces before assembly and incorporating stripe coats
was among the best performing scenarios, though the practice would benefit by adding limited
caulking to reduce water ingress.

Figure 17. Minimum, average, and maximum pits measured on the crevice face of the angle for panels
representing new steel coating systems.
For the weathered steel maintenance painting scenarios (Figure 18), the observed pitting is
probably dominated by what occurred during the pre-weathering exposure. However, by
comparison to the “nothing” set of data, there is some evidence that four scenarios result in
less pitting:
• stripe and full caulk for abrasive blasted panels
• stripe and full caulk for power tool cleaned panels
• stripe and ¾-caulk for abrasive blasted panels
• and stripe and top-caulk for power tool cleaned panels
The benefit of all caulking scenarios is corroborated by the visual observations. The panel
geometry contains 11 unique crevice surfaces. Analysis of the remaining crevice surfaces may
provide sufficient statistical data to get additional insight.

Figure 18. Minimum, average, and maximum pits measured on the crevice face of the angle for panels
representing weathered steel coating systems.
 Conclusions

The following conclusions were formed based on the results of the study:
• Galvanized bolts perform better over time than black oxide bolts.
o If black oxide bolts are utilized, proper surface preparation along with additional
stripe coats will help prevent corrosion from occurring.
• Stripe coats and caulking of crevices directly exposed to water/ moisture will help prevent
crevice corrosion on new steel.
o When caulking newly applied steel consider leaving the bottom crevice uncaulked
to allow moisture to escape.
• When working with weathered steel, full stripe coats and caulking of all crevices provided
the best results in regards to reducing crevice corrosion and pitting.
• As a best practice, mating steel surfaces should receive a primer coating prior to assembly
 Acknowledgments

This material is based upon work supported by Triborough Bridge and Tunnel Authority (New
York, NY) and performed by Elzly Technology. Any opinions, findings and conclusions or
recommendations expressed in this material are those of the author(s) and do not necessarily
reflect the views of the Triborough Bridge and Tunnel Authority.
Eric Shoyer, Elzly Technology Corporation

Eric Shoyer has worked with Elzly Technology Corporation in various aspects of corrosion
control and materials engineering for over 8 years. Shoyer is an active member of several
technical societies including SSPC and NACE International, and is a registered NACE CIP Level 2.
Shoyer has a B.S. degree in civil engineering and a B.S. degree in structural engineering both
from Drexel University.