Recently Developed Bristle Blasting Process for Corrosion ... DEVELOPED BRISTLE... · Robert J. Stango, Ph.D. ([email protected]), and Piyush Khullar . Mechanical Engineering Department

RECENTLY DEVELOPED BRISTLE BLASTING PROCESS

FOR CORROSION-REMOVAL

Robert J. Stango, Ph.D. ([email protected]), and Piyush Khullar Mechanical Engineering Department

1515 West Wisconsin Avenue Marquette University

Milwaukee, WI 53233 USA

ABSTRACT Currently, an array of surface cleaning processes and equipment such as grit blasting, needle guns, and a variety of non-woven and coated abrasive tools are used for removing corrosion from metallic surfaces. This technical paper describes a recently developed surface treatment method termed the bristle blasting process. The process uses a specially designed rotary bristle tool, which is attached to a power tool spindle that operates at approximately 2,500 rpm. Together, the bristle tool and power tool are precisely tuned in order to provide an impact and immediate retraction of bristle tips from the corroded surface. Key issues that are examined in this paper concerning the design and performance of bristle blasting processes include:

1. basic collision mechanics of the bristle tip/workpart surface 2. kinetic energy equivalence between rotary bristles and grit blast media, 3. texture/profile of surfaces that are generated by bristle blasting processes, and 4. corrosion removal performance of bristle blasting and the cleanliness of surfaces that are generated by the bristle blasting process.

Case studies are reported that assess the corrosion removal performance as well as the life expectancy of bristle blasting tools for applications involving API 5L steel, which is commonly used in the petroleum industry for onshore/offshore applications. Surfaces generated by the bristle blasting process are shown to mimic the visual cleanliness and anchor profile that is characteristic of grit blasting processes.

KEY WORDS

Anchor Profile; Bristle Blast; Corrosion Removal; Removal of Corrosive Products; Surface Cleaning; Surface Preparation.

1 For more information please contact [email protected] or visit www.processaids.in

REVIEW OF MECHANICAL SURFACE CLEANING TECHNOLOGIES

Several mechanical technologies/processes that are currently used for cleaning and texturing metallic surfaces are summarized below. In each case, these processes involve the use of specially designed tools or media that are capable of removing both surface contaminants and substrate material. Generally, the hardness of the tool/media exceeds that of the target surface and consequently, repetitious contact between the interface of the tool and workpart surface leads to cutting/erosion of both the contaminant and subsurface materials. As a final outcome, the workpart surface bears tool signature markings which, collectively, define the texture/profile of the cleaning process. Wire Brushes Wire brushing tools are comprised of flexible metallic bristles that are anchored to a rotating hub, as shown in Figure 1a. As the hub rotates, wire tips repeatedly contact the workpart surface and generate striations, or score markings throughout the region of contact. These striations are caused by bristle tips, which essentially plow through the contact zone, thereby generating a multitude of parallel troughs that remove both surface debris and parent material. Consequently, the textured surface consists of striations/score markings depicted in Figure 1b,

(a) (b)

Figure 1 (a) Conventional wire brush and (b) typical brushed surface illustrating continuous score markings generated throughout the contact zone.

which trace the path that individual bristle tips have traversed during the material removal process. Needle Guns As shown in Figure 2a, the needle gun consists of a bundle of parallel wire rods or “chisels” that are placed in contact with the workpart surface. When the tool is activated, the wires rapidly oscillate back and forth (i.e., along the axial direction) thereby causing repeated contact and indentation between the wire tips and target surface. This repeated contact, in turn, leads


(a) (b)

Figure 2 (a) 12-rod pneumatic needle gun and (b) typical textured surface after corrosion removal. to the removal of surface debris and simultaneously generates the coarsened surface texture depicted in Figure 2b. Bonded Abrasives The term bonded abrasives refers to a variety of grinding tools that use sharp granular mineral to cut through corroded surfaces. The construction/format of the tool can vary and includes a wide range of rotating/reciprocating coated abrasive pads, discs (see Figure 3a) or wheels, and non-woven abrasive pads, discs or wheels. In any case, the principle of operation

(a) (b)

Figure 3 (a) Power abrasive disc and (b) ground/treated surface illustrating score markings generated by cutting mineral throughout the contacting region.

is similar; that is, while force is exerted onto the abrasive-laden tool, the mineral cuts through both the surface contaminant and base metal. Consequently, the textured surface consists of striations/score markings (see Figure 3b) that trace the path of individual grains during the material removal process. The use of these tools is known to generate high temperatures within the substrate, which may or may not be accompanied by surface burn marks/discoloration. In any case, high surface temperatures are known to create conditions that can promote damage to the substrate and/or accelerated corrosion of the cleaned surface1-3. The application of greater force to grinding tools can be accompanied by greater surface temperature, which can


promote greater damage to the substrate. Therefore, the use of excessive tool force must be avoided, even though this may increase the rate of cleaning/worker productivity. Despite widespread publicity of this important issue in the engineering community, a cursory review of SSPC/NACE surface preparation standards by one of the authors has shown an absence of instructional or cautionary information communicated to practitioners in the surface preparation community.

Grit Blasting Grit blasting processes utilize “loose” abrasive grains, which are propelled toward the target surface, as shown in Figure 4a. Upon impact, each abrasive grain forms a “crater-like” micro-indentation, which simultaneously removes friable corrosion and exposes fresh substrate

(a) (b)

Figure 4 (a) Schematic of abrasive blasting process, and (b) characteristic grit blast surface generated by G16 steel media.

material. As depicted in Figure 4b, the repeated micro-indentations generate a surface texture that consists of “peaks and valleys”, which are associated with the recurring impact and deformation of the workpart surface.

Bristle Blasting The recently developed bristle blasting tool (see Figure 5a) has a brush-like appearance, and consists of sparsely populated wires having tips that are both hardened and sharpened. As the spindle rotates, each bristle tip strikes the metallic surface and immediately


(a) (b)

Figure 5 (a) Recently developed bristle blasting power tool system (pneumatic version shown), and (b) characteristic surface generated by bristle blasting process.

retracts/rebounds, thereby causing a multitude of impact craters that are similar to those formed during grit blasting operations. This repetitious process both removes the corrosive layer and generates a fresh surface having the coarse “shoveled” pattern shown in Figure 5b.

INTRODUCTION TO BRISTLE BLASTING PROCESS

Historical Studies in Bristle Impact Mechanics The study of bristle impact mechanics is a relatively recent topic of interest4-9. As shown in Figure 6, the problem consists of examining the dynamic response of an individual bristle or monofilament that is attached to a rotating hub and undergoes periodic contact with a rigid, flat surface. That is, as the hub rotates (counterclockwise) the bristle tip initially engages the workpart surface at s=0, and remains in contact with the surface until the release point s = sR is reached. Throughout the excursion, the successive normal forces f1, f2, …, fp that is

Figure 6 Schematic of single bristle illustrating position of the filament at initial contact (S=0), intermediate position, and final position (S=SR), and variation of normal forces exerted by bristle tip throughout the contact zone.


generated along the interface of the bristle tip/workpart surface have been calculated via analytical/numerical methods. Some key results4 are shown in Figure 7a for three different (progressively increasing) non-dimensional hub speeds ω* = 0.2, 0.5, and 1.0, and indicate that both the contact length and the magnitude of the contact forces can vary significantly

(a) (b)

Figure 7 Solution based upon analytical model of bristle dynamics, illustrating (a) variation of (non-dimensional) contact force fn throughout the contact zone at three different (non-dimensional) operating speeds, and (b) repeated contact-and-rebound phenomenon that results at highest hub rotating speed.

as the hub speed is increased. That is, slower hub speeds lead to both a greater region of contact and a reduction in the sustained peak magnitude of the normal contact force, whereas faster hub speeds lead to both a lesser region of contact and an increase in the peak magnitude of the normal contact force. Moreover, further results shown in Figure 7b indicate that if a hub speed of sufficient magnitude (ω* = 2.0) is chosen, then the bristle tip can initially strike and retract (i.e., “rebound”) from the workpart surface, and subsequently re-strike the surface during the same cycle of rotation. In summary, the foregoing analysis of the problem suggested that monofilament bristle impact can be accompanied by primary (i.e., initial) impact, and can later be followed by a series of secondary impact events. This dynamical behavior was later experimentally confirmed by the use of high-speed strobe videography10. In order to investigate any potential relevance that the above theoretical findings could have on a fully-populated brushing tool, an experimental method was later developed for measuring the actual variation of dynamic bristle forces within the contact zone11. A key outcome of this research is shown in Figure 8, whereby the measured average force exerted by bristle tips throughout the contact zone is shown for three different brush rotational speeds n1 = 550 rpm, n2 = 1,300 rpm, and n3 = 2,850 rpm. In each case, the forces exerted by the bristle tip within the initial portion of the contact zone abruptly rise and are confined to a narrower band-width


Figure 8 Actual (i.e., measured) force gradient generated within contact zone of fully populated brush (bristles moving from right to left) at three different hub speeds, indicating abrupt rise in bristle contact forces as filament enter the initial region of contact.

as the rotational speed of the hub is increased. These results indicate that bristle tips do indeed exert transient forces of significant magnitude near the point of entry within the contact zone. However, one may observe that the bristle tips remain in contact with the surface throughout the entire excursion, because the retraction/rebound of bristle tips is suppressed due to the presence of tightly-packed adjacent filaments. Consequently, brushing processes are known to generate score markings or striations along the entire region of contact. More recently, the feasibility of devising a bristle geometry that exhibits a single, primary impact crater was explored for potential use in bristle peening tool applications12. To this end, the dynamic response of three differently shaped bristles shown in Figure 9 were examined, namely, a reverse-bent bristle (9a), straight bristle (9b) and forward-bent bristle (9c). In each case, the bristle rotates in the counterclockwise direction at approximately 2,500 rpm, and the

(a) (b) (c) (d)

Figure 9 Geometry of three differently configured rotating bristles, featuring 9a (reverse bent knee), 9b (straight without bend) and 9c (forward bent knee). Upon impact with the workpart surface (bristle is moving from left to right), the bristle tip rebounds, and its location is precisely monitored via high-speed digital camera and shown in 9d.

precise position of the spherical tip (monitored via high-speed digital camera) is depicted in Figure 9d throughout the pre-and post-collision period. Careful examination of the bristle tip trajectory for all three cases shows that the initial impact of the bristle tip is followed by


retraction/rebound from the surface, and is not accompanied by a secondary impact. The dynamic response of the forward-bent bristle is of greatest interest, and indicates that no sliding motion occurs between the tip/workpart surface throughout the duration of contact. This finding plays a key role in modifying the dynamic behavior of bristle tools and can have important consequences on surface cleaning and finishing processes. Development of Bristle Blasting Tool Dynamic properties of the forward-bent bristle can be effectively used for generating a single/primary impact crater that mimics the surface morphology ordinarily associated with grit blasting processes. To this end, the sequence of several consecutive frames acquired by a high-speed digital camera is shown in Figure 10 for a bristle rotating in the counterclockwise

Figure 10 Successive frames of a single bristle taken from high-speed digital camera depicting the approach (frames 1, 2, and 3), contact/collision (frame 4), subsequent retraction (frame 5), and return to equilibrium position (frames 6-11) of bristle.

direction at 2,500 rpm. As the bristle tip approaches the workpart surface (motion is from left-to right) initial contact is made at the indicated point of impact. Upon striking the surface, a crater-like micro-indentation is formed, and the bristle tip subsequently rebounds from the surface. Throughout this duration the hub continues to rotate and the final trajectory of the bristle tip results in a single/primary impact site. Typical impact craters that are formed on a ductile steel surface (API 5L) are shown in Figure 11, and have been likened to shoveling craters that are commonly generated by grit blast media13.

Figure 11 Typical impact craters generated by bristle blasting tool (material system: API 5L)


Details concerning the design of a bristle blasting tool are shown in Figure 12, whereby sparsely populated bristles are attached/protrude through the hub or belt (Figure 12a), which is

(a) (b) Figure 12 (a) Design and construction of the bristle blasting tool, and (b) functional regions of the bristle, including the bristle tip, shank, knee, and main body.

constructed from a flexible, high-strength, fiber-reinforced polymer that both dissipates and stores energy during the collision process. In addition, bristle tips are hardened (Rc ≈ 65) and sharpened (Figure 12b), thereby providing optimal corrosion removal performance. Finally, a rigid plastic mounting ring is fitted to the internal diameter of the belt, and this assembly is subsequently secured to the die-cast aluminum hub that is attached to the drive spindle.

Kinetic Energy Equivalence for Grit Blast Media and Bristle Blast Wire The kinetic energy content of a grit blast particle is regarded as an important measure for assessing the capacity that blast media has for both removing corrosive layers and forming micro-indentations that are needed for achieving a requisite anchor profile14. In this section, both the kinetic energy of a grit particle and a rotating bristle are computed; this, in turn, provides a foundation for comparing the relative performance that one may expect when using the two different processes. Kinetic energy of blast media A schematic representation of the grit blasting process is shown in Figure 4a and consists of a pressurized system that ejects media from a nozzle at speeds that typically range from 30-120 m/s. The kinetic energy of grit particle ep is customarily computed15

ep = ½ mpvp2 sin2

α (1)

where vp is the speed of a grit particle having mass mp, whose supply nozzle is inclined at an angle α relative to the horizontal target surface. Kinetic energy of a standard rotating bristle and “energy equivalence” Next, the kinetic energy of a rotating bristle eb is computed for a wire bristle that rotates under “standard conditions”, i.e., at constant spindle speed, n (rpm). Based upon rigid body mechanics, the bristle energy is thus computed16


eb = ½mbvG2 + ½IGω2 (2)

where vG is the speed of the bristle mass center, IG is the mass moment of inertia of the bristle about the mass center G, and ω is the angular velocity of the spindle, which is directly related to the spindle speed n (rpm) via the relation ω = nπ/30. A direct comparison can now be made between the kinetic energy of blast media and the corresponding speed of a rotating bristle having the equivalent kinetic energy. Thus, direct equality of the two different media energies in Equation (1) and Equation (2) yields

1

2

22

sin

)2

(121

30

−

⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢

⎣

⎡++

=απ

rLL

mmv

np

bp (3)

where L is the nominal bristle length, and r is the radius of the bristle tool hub (reference Figure 12). Equation (3) can be used for comparing the “equivalent” energy content of the two different surface cleaning processes as follows:

Step 1: Select/identify the grit blast process parameters Referring to the right hand side of Equation (3), the selection of grit media (i.e., grit material composition and mesh size establishes the particle mass mp; vp corresponds to the estimated nozzle exit speed of media; α is the nominal nozzle angle of orientation that is used during the grit blast operation (typically, 70 deg.). Step 2: Select/identify the geometry of the bristle blasting tool Referring again to the right hand side of Equation (3), the selection of nominal steel wire length L and bristle diameter establishes bristle mass mb; r corresponds to the hub radius of the bristle blasting tool. Step 3: Compute the corresponding rotational speed of the bristle blasting tool Finally, with all terms defined on the right hand side of Equation (3), the rotational speed n (rpm) of the bristle blasting tool spindle can be computed that corresponds to the “equivalent” kinetic energy that is generated by the use of standard grit blast process parameters.

The usefulness of Equation (3) is shown in Figure 13 for several different standard steel media mesh sizes (G12, G16, …, G25) in conjunction with the dimensions of a commercially

Figure 13 Relationship between spindle speed (standard bristle motion) and grit velocity for various steel media sizes. (Note: spindle speed 2600 rpm corresponds to grit velocity of 35 m/s for G16 steel media)


available bristle blasting tool shown in Figure 12. As a practical illustration, for example, the use of G16 media (diameter ≈ 1mm) having a nozzle exit speed of 35 m/s corresponds to bristle blasting tool operating at the spindle speed n = 2,600 rpm. Kinetic energy of enhanced bristle motion. Next, an alternate design of the bristle blasting power tool is examined that will enhance the kinetic energy of a wire bristle without increasing the spindle speed of the tool. The concept is shown in Figure 14a, and involves the use of a stationary obstruction (termed an accelerator bar) that is strategically placed in the path of an oncoming, rotating bristle. Thus, upon striking the cylindrical bar, the bristle tip retracts and stores additional (potential) energy prior to being released (Figure 14b). Upon recoil, the potential energy is converted to kinetic energy and the bristle acquires additional speed prior to impact with the target surface. With the aid of a high-speed digital camera, this aspect of the problem has been recently examined by the authors16.

(a) (b)

Figure 14 (a) Depiction of bristle tips initial contact with the accelerator bar and subsequent rear-ward retraction, and (b) acceleration of bristle tip towards the target surface upon release from the accelerator bar.

The results of their work (details presented in Ref. 16) are summarized below, and yield the following revised relationship between grit blast speed and rotational speed of the bristle blasting tool:

{ }21sin1 AA

mmv

p

bp +=

α (4)

with

A1 = 2

302⎟⎠⎞

⎜⎝⎛ + hrnKL π , (5)

A2 = ( 2112

L K ) , (6)

and K = 1208.5. The potential benefits offered by the accelerator bar can now be assessed by using Eqs.(4-6) to re-construct Figure 13. Thus, the results are shown in Figure 15, and


indicate that a considerable gain can be made when compared to standard bristle motion. For example, the use of G16 media is re-examined, and indicates that a bristle blasting tool

Figure 15 Relationship between spindle speed (including enhanced bristle motion) and grit velocity for various steel media. (Note: spindle speed 2600 rpm corresponds to grit velocity of 79 m/s for G16 media).

operating at the same spindle speed (i.e., n = 2,600 rpm) now corresponds to a grit nozzle exit speed of 79 m/s. A direct comparison of this result with standard bristle motion (i.e., 35 m/s) indicates that a grit speed ratio of 2.3 is gained by introducing the accelerator bar into the bristle blasting tool system.

DESCRIPTION AND IMPLEMENTATION OF BRISTLE BLASTING PROCESS Overall appearance and select features of the bristle blasting tool system are shown in Figure 16 for the pneumatic version of the power tool. The power tool itself consists of the main body, control handle, accelerator bar, dust collector, and a spindle that operates at approximately 2,500 rpm. The bristle blasting tool/hub assembly directly attaches to the

Figure 16 Overall description and key features of bristle blasting system. spindle, and both the main body and control handle are used for securing/holding the power tool during operation. Details concerning skillful use of the tool for corrosion removal


applications are discussed in the following section. Proper safety precautions involving protective wear and safe practice, however, are not provided in this paper.

Standard/Recommended Implementation of the Bristle Blasting Tool All manual surface treatment processes require dexterity, visual acuity, and a basic understanding of key parameters that affect the performance of surface finishing equipment. Training and experience are, therefore, important factors that enable users to develop skills that are needed for a successful outcome. The skill-sets that are essential for successful application of the bristle blasting process are quite similar to those needed for other surface treatment processes, and include the following: 1) proper orientation of the tool in relation to the target surface, 2) control of tool force exerted onto the surface, and 3) the feed rate and direction of the tool during operation. In the following discussion, each of these user-based considerations is briefly discussed within the context of a common corrosion removal application. Initializing the process cleaning parameters Appropriate selection of the bristle blasting process parameters can be readily established by first, identifying a candidate surface that requires cleaning, and isolating a portion of the surface for initial cleaning/testing. In general, the face of the tool hub is oriented perpendicular to the treated surface during use, as shown in Fig. 17. During corrosion removal, the bristle tips are brought into direct contact with the corroded surface using minimal applied force, and the rotating tool is gradually moved along the transverse direction, that is, either to the left or right of the user (see Fig. 17a). Thus, the appropriate pressure and feed rate of the tool is obtained by direct experimentation and by visually inspecting the trial-tested region to ensure that the desired cleaning standard/requirement is reached. Method/pattern for continuous systematic cleaning Having obtained the appropriate process parameters for corrosion removal, the user then identifies the region to be treated, and develops a simple plan for obtaining complete coverage. As shown in Fig. 17a, for example, the surface of a corroded steel component must be cleaned. The user, in turn, has elected to begin the corrosion removal process at the

(a) (b) (c)

Figure 17 Recommended use of bristle blasting tool for corrosion removal. First, a horizontal row is prepared (Fig. 17(a)) using minimal applied force and steady feed rate. The process is then repeated by overlapping the second row (Fig. 17(b)) with the previous row that was cleaned. Finally, the entire surface is cleaned (Fig. 17(c)) by repeatedly overlapping each row with the previously cleaned region until full coverage is completed.


extreme left end of the component, and has applied the working surface of the tool along the transverse direction, i.e., from left to right. This procedure has resulted in a cleaned and textured horizontal band or row, which appears in Fig. 17a. Equally important, the user has started the cleaning operation along the top (uppermost) portion of the corroded surface, and will perform all subsequent cleaning by the use of overlapping bands that have their starting point below (under) the previously cleaned region. That is, correct use and optimal cleaning/texturing performance of the tool requires that each overlapping successive band is generated beneath the previously cleaned region/row. Therefore, as shown in Fig.17b, the user has correctly overlapped the previously cleaned region, and has generated/cleaned the next row by placing the working surface of the rotating tool directly below the initially prepared surface. Completing the corrosion removal process Corroded components can be completely cleaned by repeating the previously described procedure. Thus, as shown in Fig. 17c, the top surface of the corroded beam has been completely cleaned, and the user is ready to remove corrosion from any remaining surfaces. Finally, if any portion of the surface is identified where unsatisfactory cleaning has been obtained, the user can return to these locations for final “touch-up” cleaning, as needed.

PERFORMANCE OF BRISTLE BLASTING TOOL

Corrosion Removal: Surface Cleanliness and Morphology

Performance of the bristle blasting process is now examined within the context cleaning and profiling the API 5L piping shown in Figure 18. Careful examination of the corroded surface suggests that SSPC Condition D (100% rust with pits) accurately assesses the initial severity of corrosion present.

Figure 18 Corroded section of 6 in. diameter pipe (API 5L) used for evaluating corrosion removal performance of bristle blasting tool.


In Figure 19 an interior section of the cleaned specimen is shown after bristle blasting, and an initially corroded specimen is place directly below for comparison. Further examination of the surfaces appearing in Figure 19 is shown in Figures 20a and 20b at low magnification

Figure 19 Interior of API 5L piping in as-received condition (bottom) and after bristle blast cleaning (top).

using an optical microscope. Inspection of Figure 20a indicates that while a significant depth of rust and corrosive pits appear on the initial specimen, the bristle blasted surface (Figure 20b) has a uniform, corrosion-free appearance, and no corroded pits remain on the cleaned specimen. In Figure 21, a scanning electron microscopic (SEM) image of the cleaned surface shows both the exposed substrate metal and the detailed surface texture. Careful examination

(a) (b)

Figure 20 (a) Photograph depicting the extent of corrosion/pitting on the as-received surface of piping, and (b) cleanliness of the bristle blasted surface after corrosion removal.

of this image (100X) reveals that the surface is free of residual corrosion and that the characteristic impact crater appearing in Figure 11 (i.e., shovel micro-indentation) appears as a repeated pattern along the cleaned surface.

Figure 21 Scanning electron micrograph (magnification 100x) of the treated surface shown in Fig. 20(b)


A direct comparison can now be made between the cleanliness of surfaces generated by the bristle blasting process with those published by the Society for Protective Coatings (SSPC) (formerly, Steel Structures Painting Council) for power hand tools (SSPC VIS 3)17. Such a comparison clearly indicates that the surfaces generated by the bristle blasting process surpass the cleanliness that is characteristic of all power tool cleaning processes, including hand tool cleaning by power brushes, sanding discs, and needle guns. Cleanliness of the surfaces produced by bristle blasting also outperforms the cleanliness and texture expectations that are typical of power tool cleaning to bare metal, as cited in SSPC standard SP 11. That is, SP 11 allows corrosion to remain at the bottom of pits and has a minimum surface profile requirement of 25 microns, whereas no corrosive pits remain after bristle blasting, and the surface profile typically varies from 63 to 84 microns, as demonstrated in the next section. Comparison can also be made between the bristle blasting process and the dry abrasive blast cleaning standards, namely SSPC VIS 118. Careful examination of SSPC photographs for these visual standards indicates that the cleanliness performance of the bristle blasting process exceeds that of brush-off blast cleaning (SP 7), industrial blast cleaning (SP 14), and commercial blast cleaning (SP 6). The thoroughness of the bristle blasting process, however, does appear to be comparable with near-white blast cleaning (SP 10) and white metal blast cleaning (SP 5).

Material Removal Studies

The removal of corrosive layers via mechanical surface treatment processes is inevitably accompanied by the removal of base material as well. Excessive removal of substrate material, however, can compromise the integrity of the component/structure, especially in regions where thin cross-sections may exist. Therefore, considerable experimentation has been carried out to assess the material removal performance of the bristle blasting tool, and a portion of these results are shown in Figure 22 and Figure 23. The material removal process was carried out by penetrating the rotating tool into a flat, ground surface (API 5L) at a specific penetration depth, and then allowing the tool to extract parent material for a pre-defined time interval without interruption. After each interval, the specimen was weighed using a high

Figure 22 Measured material removal rate for API 5L flat specimen, using bristle blasting tool having a 25 minute duty cycle at different depths of penetration. Approximate bristle tool specifications: face width: 22


mm, hub radius: 27.5 mm, bristle wire diameter: 0.73 mm, bristle length: 27 mm, total bristle population ~480.

resolution electronic balance, and the difference in weight of the specimen (equal to the material removed) was recorded. In Figure 22 results are shown for the material removed at three different penetration depths by a tool that has been “aged” for 25 minutes, that is, has been previously subjected to a duty cycle consisting of 25 minutes of continuous service. Thus, the material removed at a penetration depth of 1.5 mm (diamond), 4.0 mm (square) and 7.5 mm (triangle) are reported and indicate that the material removal capacity of the tool increases with increasing depth of penetration. A similar experiment was carried out in order to assess the role that the age of the tool plays in material removal performance. These results appear in Figure 23, where the

Figure 23 Measured material removal rate for API 5L flat specimen, using bristle blasting tools having various periods of continuous service. Approximate bristle tool specifications: face width: 22 mm, hub radius: 27.5 mm, bristle wire diameter: 0.73 mm, bristle length: 27 mm, total bristle population ~480.

material removed from a specimen by tools that have acquired three different duty cycles, namely 5 minutes (triangle), 25 minutes (square) and 72 minutes (diamond) of service is shown. In each case, the experiment was performed at a nominal penetration depth of 4mm. The results suggest that as the period of tool use increases, bristle tips are prone to eventual wear and/or breakage, which progressively reduces the material removal capacity of the tool. Surface Texture and Assessment of Tool Life

The use of various tools and media for surface cleaning processes inevitably yields a characteristic surface appearance or “signature” that is peculiar to the contact/interaction that occurs along the interface of the tool and workpart However, finer details of the surface morphology cannot be visually resolved and, therefore, a host of surface texture parameters have been proposed for quantifying architectural surface features with greater precision. To this end, the average peak-to-valley texture parameter Rz is often used for measuring the “anchor profile” of cleaned surfaces prior to the application of paints and coatings. In order to develop an understanding of the detailed surface texture that is generated by the bristle blasting process, surfaces are prepared by using the standard procedure previously


discussed and shown in Figure 17a. Accordingly, the contact zone shown in Figure 24a has been generated using the platform of a three-axis milling machine by traversing the surface of a flat, ground API 5L steel specimen at a constant lateral feed rate of 0.5 cm/s. Surface texture parameter Rz was subsequently measured using standard laboratory equipment

(a) (b)

Figure 24 (a) Typical specimen prepared by single pass of bristle blasting tool, and (b) measured surface profile at several locations in the contact region where s=constant. These data show the role that increased penetration depth plays in generating surface profile.

(Mitutuyo Surface Roughness Tester SurfTest SJ-301), and is reported in Figure 24b at several uniformly-spaced sampling positions that are coincident with the transverse coordinate (s=constant). The results shown in Figure 24b have been obtained using three different tool penetration depths, namely, 0.05 in. (square), 0.10 in. (circle), and 0.25 in. (triangle), and indicate that the length of the contact zone broadens as the penetration depth increases, which is the expected result. The results also indicate that coarse texture is sustained over a wider region within the contact zone at shallower penetration, and abruptly declines at the deepest

Figure 25 (left) Bristle blasted surface generated within region S=0.2 in. of specimen shown in Figure 24(a), and (right) topographic profile of a bristle blasted surface traced by a surface profilometer.

penetration. Finally, the actual surface profile that was recorded along the coordinate position s = 0.2 in. is shown in Figure 25, which illustrates the topographical nature of a bristle blast surface.


As one may expect, the texture generated by bristle blasting tools will vary as the duty cycle of the tool increases due to filament tip wear and/or breakage. This aspect of tool performance has been examined by manually cleaning corroded API 5L pipe (Figure 18) and periodically measuring Rz using standard press-film replica tape. Thus, the relationship between duty cycle and profile performance is shown in Figure 26 and indicates that a new tool (i.e., as-received) generates an average surface profile of Rz ≈ 84. As the duty cycle of the tool increases, however, the surface texture regularly declines and, after approximately one hour of service, is reduced to Rz ≈ 62.

Figure 26: Variation of surface texture/anchor profile as bristle blasting tool progressively ages. Approximate bristle tool specifications: face width: 22 mm, hub radius: 27.5 mm, bristle wire diameter: 0.73 mm, bristle length: 27 mm, total bristle population ~480.

Heretofore it has been observed that both the material removal capacity and texture parameter Rz decline as the duty cycle of the bristle blasting tool increases. In order to gain insight into the probable cause of this erosion of performance, SEM images of the bristle tip have been examined at various stages of tool use. In Figure 27a the appearance of a sharpened, as-received bristle tip is shown. A typical bristle tip is once again shown in Figure 27b after 25 minutes of continuous service and indicates that significant erosion of the bristle tip has occurred as well as wear along the lateral surface where contact is made with adjacent filaments. Finally, the bristle tip shown in Figure 27c has been subjected to 60 minutes of continuous service, and indicates that extensive erosion and damage/fracture has been sustained by the bristle tip, along with further thinning of the shank due to interbristle contact. Although continued use of the tool beyond 60 minutes is feasible, further damage/breakage of bristle tips is inevitable, and retirement of the tool is eminent.


Bristle Blasting vs. Grit Blasting: Comparative Studies

Although little comparative data can be found in the literature for bristle blast/grit blast processes, some new and recent finding are presented in this section. The initial surface of a corroded steel specimen that was used in this portion of the study is shown in Figure 28. This specimen was subsequently cleaned via bristle blasting using a tool that had acquired fewer

(a) (b) (c)

Figure 27 Scanning Electron Microscopic (SEM) images of bristle tips having different duty cycles, namely 27a (0 minutes (as-received)), 27b (25 minutes), and 27c (72 minutes).

than 5 minutes of continuous service. Similarly, the corroded surface was cleaned under controled conditions via grit blasting by using (as-received) G16 steel grit in conjunction with a nozzle diameter of 1.2 cm, grit exit speed 170 m/s, stand-off distance d = 45 cm, and

Figure 28 Typical initial surface of corroded steel specimens used in comparative studies of adhesive bond strength for surfaces prepared by bristle blasting and grit blasting processes.

approximate nozzle angle α = 70 deg. The general appearance of bristle blasted and grit blasted surfaces are shown in Figures 29a and 29b, respectively, and is accompanied by the


measured characteristic profile for each surface. In each case, the cleaned surfaces have a coarse, corrosion-free appearance, with no evidence of any remaining corrosive pits. Measurement of the bristle blast surface yielded Rz = 102 microns along with the average peak count RPc/cm = 12, whereas the grit blast surface recorded Rz = 133 microns along with RPc/cm = 10. Performance of the two differently prepared surfaces was assessed by measuring the adhesive bond strength of a two-part epoxy coating system (Pratt-Lambert epoxy primer S3393 & S3344, and epoxy S3400 & S3498). Although some debate exists concerning the

(a) (b)

Figure 29 General appearance of cleaned surface and corresponding measured profile for (a) bristle blasted surface and (b) grit blasted surface. interpretaion of this standard test, the coating adhesion test (ASTM D 4541) is often regarded as a key indication of the integrity of both the prepared surface and the coating material system. Each of the specimen surfaces was prepared, coated, and stored in a dessicator for 60 days. Subsequently, standard dollys were bonded to the coating, and the interfacial bond


Figure 30 Result of coating adhesion test for surfaces prepared by bristle blasting process (triangle) and grit blasting process (circle).

strength (MPa) was measured using standard laboratory apparatus (Defelsco PosiTest Pull-Off Adhesion Tester). Result of the adhesion test is shown in Figure 30 and shows near-parity for the two different processes. That is, the coating bond strength of the bristle blasted surface meets or exceeds the bond strength of surfaces prepared by the grit blasting process. Also, the typical condition of the de-bond surface for each steel specimen is shown in Figure 31(a) for the bristle blast surface and in Figure 31b for the grit blast surface. In each case, the specimen surfaces appears similar, and failure has occured along the coating/steel interface.

(a) (b)

Figure 31 General appearance of the de-bond/interface surfaces for steel components prepared by (a) bristle blast and (b) grit blast processes. These above results are now compared with recently published experimental data19 for several different coating systems, which has been reproduced in Figure 32. As indicated by the legend in Figure 32, surfaces were prepared by the bristle blasting (MBX), grit blasting


Figure 32 Recently published coating adhesion strength test data reported in Ref. [19] for specimens cleaned via grit blast, bristle blast, and power wire brush processes.

(GRT), and power wire brushing (PWB) processes. As in the present study, the data reported in Ref [19] indicate near-parity for coating adhesion performance of bristle blast surfaces and grit blast surface. In each case, however, a reduction of coating bond strength has been obtained for power wire brushed surfaces.

SUMMARY AND CONCLUSION

Bristle blasting is a viable, aggressive process for removing corrosive layers while simultaneously generating an anchor profile. The corrosion removal capacity, surface cleanliness, and coating adhesion performance of the bristle blasting process is on an equal par with norms and standards that are commonly associated with grit blasting operations. This conclusion has been corroborated by both independent laboratory tests and in-field service experiences, which are currently being compiled and assessed by one of the authors. A direct comparison of bristle blasting tool performance with SSPC VIS 3 in conjunction with the power hand tool cleanliness standard SP 11 clearly indicates that the bristle blasting process surpasses the cleanliness and profile that is characteristic of all power tool cleaning processes, including hand tool cleaning by power brushes, sanding discs, and needle guns. Furthermore, comparison of the bristle blasting process with dry abrasive blast cleaning standard SSPC VIS 1 shows that performance of the bristle blasting process exceeds that of brush-off blast cleaning (SP 7), industrial blast cleaning (SP 14), and commercial blast cleaning (SP 6). Thoroughness of the bristle blasting process, however, appears to be on an equal par with near-white blast cleaning (SP 10) and white metal blast cleaning (SP 5). Thus, the disparity of the bristle blasting process with norms/standards cited in SP 11 suggests that a re-evaluation of this document is needed in order to accurately convey the performance of bristle blasting processes to the corrosion/surface finishing community.

ACKNOWLEDGEMENT

The authors gratefully acknowledge support provided by the project sponsor, Monti Werkzeuge, Bonn, Germany, and the technical advice offered by Mr. Werner Montabaur.


Also, the expert assistance of Raymond Fournelle, Professor of Mechanical Engineering, in preparing the Scanning Electron Microscope (SEM) images is deeply appreciated.

REFERENCES

1. Chen, X., Rowe, W.B., and McCormack, D. F., 2000, Analysis of the Transitional Temperature for Tensile Residual Stress In Grinding, Journal of Materials Processing Technology, 107, 216-221.

2. Xiao, G., Stevenson, R., Hanna, I. M., and Hucker, S. A., 2002, Modeling of Residual Stess in Grinding of Nodular Cast Iron, Journal of Manufacturing Science and Engineering, 124, 833-839.

3. Kruszynski, B. W., and Wojcik, R., 2001, Residual stress in Grinding, Journal of Materials Processing Technology, 109, 254-257.

4. Shia, C. Y., Stango, R.J., and Heinrich, S.M., 1993, Analysis of Contact Mechanics for Circular Filamentary Brush/Workpart System-Part I: Modeling and Formulation, Proceedings of the ASME Symposium on Contact Problems and Surface Interactions in Manufacturing and Tribological Systems, 67 (4), 171-179.

5. Shia, C. Y., Stango, R.J., and Heinrich, S.M., 1993, Analysis of Contact Mechanics for Circular Filamentary Brush/Workpart System-Part II: Solution Method and Numerical Studies, Proceedings of the ASME Symposium on Contact Problems and Surface Interactions in Manufacturing and Tribological Systems, 67 (4), 181-190.

6. Hatman, V., G., Haque, I., and Bagchi, A., 1996, Dynamics of a Flexible Rotating Beam Interacting with a Flat Rigid Surface, Part I: Model Development, Journal of Sound and Vibration, 194 (5), 653-669.

7. Hatman, V., G., Haque, I., and Bagchi, A., 1996, Dynamics of a Flexible Rotating Beam Interacting with a Flat Rigid Surface, Part II: Numerical Solution, Journal of Sound and Vibration, 194 (5), 671-683.

8. Shia, C. Y., Stango, R.J., and Heinrich, S.M., 1998, Analysis of Contact Mechanics for Circular Filamentary Brush/Workpart System, ASME Journal of Manufacturing Science and Engineering, 120, 715-721.

9. Useche, L. V., Wahab, M. M., and Parker, G. A., 2007, Dynamics of Unconstrained Oscillatory Flicking Brush for Road Sweeping, Journal of Sound and Vibration, 307, 778-801.

10. Stango, R. J., 1993, Unpublished Research Data, Deburring and Surface Finishing Research Laboratory, Marquette University, Milwaukee, WI 53233

11. Stango, R. J., Cariapa, V., and Zuzanski, M., 2005, Contact Zone Force Profile and Machining Performance of Filamentary Brush, ASME Journal of Manufacturing Science and Engineering, 127, 217-226.

12. Wojnar, N., 2006, Design and Application of Rotary Bristle Brush for Peening Applications, M.S. Thesis, Marquette University, Milwaukee, WI 53233.

13. Stango, R. J., and Khullar, P., 2008, Introduction to the Bristle Blasting Process for Simultaneous Corrosion Removal/Anchor Profile, ACA Journal of Corrosion and Materials 33 (5), 26-31.

14. Jones, J. R., and Gardos, J.M., 1971, An Investigation of Abrasive Blasting, Journal of ASLE, 6, 393-399.

15. Budinski, K. G., and Chin, H, 1983, Surface Alteration in Abrasive Blasting, Wear of Materials, 311-318.


16. Stango, R. J., and Khullar, P., 2009, Fundamentals of Bristle Blasting Process for Removing Corrosive Layer, NACE Corrosion Conference, Atlanta GA., 2009, Paper no. 09191.

17. SSPC-VIS 3, Visual Standard for Power- and Hand-Cleaned Steel, Steel Structures Painting Council, Pittsburgh, PA 15213-3724.

18. SSPC-VIS 1, Guide and Reference Photographs for Steel Surfaces Prepared by Dry Abrasive Blast Cleaning, Steel Structures Painting Council, Pittsburgh, PA 15213-3724.

19. Wilds, N., 2009, Bristle Blasting as a Method for Surface Preparation, International Paint, Inc., NACE Corrosion Conference, Atlanta, GA, Paper no. 09004.


Documents

Recently Developed Bristle Blasting Process for Corrosion ... DEVELOPED BRISTLE... · Robert J. Stango, Ph.D. ([email protected]), and Piyush Khullar . Mechanical Engineering Department