Steel foam material processing, properties, and potential structural applications

NSF GRANT # 1000334 NSF PROGRAM NAME: Structural Materials and Mechanics Steel foam material processing, properties, and potential structural applica...
Author: Abigayle Conley
1 downloads 0 Views 1MB Size
NSF GRANT # 1000334 NSF PROGRAM NAME: Structural Materials and Mechanics

Steel foam material processing, properties, and potential structural applications Sanjay R. Arwade University of Massachusetts, Amherst Jerome F. Hajjar Northeastern University Benjamin W. Schafer Johns Hopkins University Mohammadreza Moradi University of Massacusetts, Amherst Brooks H. Smith University of Massachusetts, Amherst Stefan Szyniszewski Johns Hopkins University Abstract: Steel foam is a material that can now be produced at the laboratory scale using a variety of different processes that create materials with a variety of different morphologies. Steel foam has not, however, been adopted in structural applications. In this paper we review some of the methods available for processing steel foams and the material properties that result from those processes, and demonstrate a possible application of steel foam in mitigating instability in structural members susceptible to local instability. 1. Introduction: Foam and cellular materials have been produced from base materials that include polymers, ceramics, and metals such as titanium, aluminum, and copper, and such foams have been applied to solve engineering problems primarily in the aerospace, automotive, and process control domains. Steel is one of the most widely used engineering materials, yet today no foam using steel as the base material is commercially available. Perhaps because of the lack of commercial availability of steel foam, no applications have been developed or widely implemented. Research conducted over approximately the last 10-15 years has shown that it is possible to fabricate steel foams at the laboratory scale and that these foams can be made to have potentially desirable mechanical properties. Despite these substantial

advances in the materials science of steel foams, a commercially available product remains elusive, and therefore structural designers have not begun to explore the potential benefits of using steel foam in civil structural applications. To date, the only experimental investigations of the potential use of steel foam in structural applications, as opposed to material characterization tests, have been to test some one foot long steel foam filled tubes [9] and some 40mm long steel foam beams [4] to failure. The dual purposes of our research project are: (1) to experimentally characterize steel foams with respect to their cyclic, tensile, and shear response, properties that are critical to structural performance but are essentially unknown for steel foams; (2) to develop and computationally test candidate applications of steel foam that will improve the performance of civil structures by, for example, improving energy dissipation or mitigating local structural instabilities. In this paper, which represents the first six months of progress in the project, we first present a brief review of the current state-of-the art with respect to steel foam manufacturing and processing. Next, a review of the mechanical properties of steel foams produced by each of the processing methods. Finally, preliminary results that indicate that deploying steel foam in thin-walled structural members has the

Proceedings of 2011 NSF Engineering Research and Innovation Conference, Atlanta, Georgia

Grant #1000334

potential to significantly improve resistance to local instability or buckling. 2. Steel foam manufacturing processes: Table 1 lists the manufacturing processes that have been demonstrated to date for making steel foams. All of the materials listed have been produced only in small

batched under laboratory, rather than commercial, conditions. As can been seen from the included images, a wide range of cell morphologies is possible using the different processing methods, and foams with either open or closed cells are possible. Investigators have succeeded in fabricating steel foams with relative

Table 1: Steel foam fabrication processes, material design variables, relative densities, and cell morphology. Process

Powder metallurgical

Typical Image

Primary Variables

Granular foaming agents (e.g. MgCO3, CaCO3, SrCO3), cooling patterns

Min Density

0.04

Max Density

0.65

Open/Closed Cells

References

Closed

[12]

Injection molding with glass balls

Types of glass (e.g. IM30K, S60HS)

0.48

0.66

Closed

[18]

Oxide ceramic foam precursor

Ceramic / cement precursor materials

Consolidation of hollow spheres

Methods for sphere manufacture (e.g. powder-coated styrofoam, gas blowing), method of connecting spheres (e.g. sintering, glue, metal filler)

0.04

Types of working before sintering (e.g. compacted, rod length), filler materials

0.05

0.13

0.23

Open

[16,17]

Working and sintering of bimaterial rods

0.21

Either

[5,10]

0.95

Open

[14]

Composite PM / hollow spheres

Matrix material used, casting may be done instead of PM

0.32

0.43

Closed

[15] SlipReactionFoam Sintering

Dispersant, bubbling agent, and quantity of dispersant and bubbling agent used

0.12

Polymer foam precursor Polymer material used 0.04 Proceedings of 2011 NSF Engineering Research and Innovation Conference, Atlanta, Georgia

0.41

Open [2]

0.11

Open

Grant #1000334 [1]

densities (ρ = ρf/ρs) that range from 0.04 to 0.95 times the density of solid steel. In the definition of relative density ρ, ρf is the mass density of the foam and ρs is the mass density of the base material. Metal foams currently available that use aluminum, titanium, or copper as a base metals have relative densities in the range 0.05 to 0.20, and have typically been used in applications in which very high ratios of the material stiffness to weight or compressive energy absorption to weight are desired. A feature of such low density foams is that they have very low material strength relative to the base metal, with yield stress as low as 1% of the base material yield stress at ρ = 0.08 [7]. In structural applications we expect the maintenance of reasonable material strength to be critical to the satisfactory performance of the material, and therefore call particular attention the ability to achieve foams with relative density greater than 0.40 using the powder metallurgy and composite hollow sphere methods. Although high relative density is also achievable using injection molding with glass balls, the use of expensive and fragile glass balls in the fabrication process will likely restrict the potential commercialization of that particular material, at least for civil engineering applications. The working and sintering of bimaterial rods can also produce high relative densities, but creates an open cell morphology which is likely to be disadvantageous in terms of corrosion resistance. Furthermore, that process creates materials with strong anisotropies in the material properties, a feature that structural engineers prefer to avoid. Finally, although the sintering of hollow steel spheres can produce materials with relative density only up to 0.20, it is the most widely investigated fabrication procedure, and is likely the closest to commercialization. We therefore intend to focus our investigations on the application of steel foam to civil structures on materials produced by powder metallurgy, sintering of hollow steel spheres, or

composite powder metallurgy / hollow steel sphere processes. 3. Mechanical properties of steel foam: Foam materials have typically been employed in mechanical or aerospace applications in which they were asked to undergo large compressive deformations at relatively low stress, or provide substantial stiffness at extremely low weight. For that reason, characterization of steel foam material properties has focused exclusively on compression testing of small rectilinear prisms of material providing the elastic modulus and compressive yield stress of the material. Table 2 summarizes the published literature on the mechanical properties of steel foam fabricated by the various processes described in Section 2 and Table 2. In all cases, the number of experiments reported is small, usually in the single digits. This reflects the substantial challenges and costs still associated with the production of steel foam. Table 2 reports values of the elastic modulus and compressive yield stress. Several of the papers also report material hardness, which is of little consequence for civil engineering design, and there exists only one published report of the tensile capacity of a steel foam, which states tensile yield stresses on the order of 1-10 MPa. We could find no published reports on the cyclic or shear response of steel foams, and both cyclic and shear loading commonly arise in civil structural applications. Steel foams with low relative density have yield stresses on the order or 1% of typical yield stress values for bulk steel, whereas when the relative density is closer to 0.50, steel foam yield stress of up to roughly 50% of steel yield stress are achievable. These findings highlight the potentially critical role that high density foams might play in civil engineering design. Steel foam elastic moduli vary from less than 1% of the bulk property to as much as 5% of the bulk property. The are low material stiffnesses, and point out that

Table 2: Summary of experimental characterization of steel foam material properties

Proceedings of 2011 NSF Engineering Research and Innovation Conference, Atlanta, Georgia

Grant #1000334





σs

= Aσc ρ 3/ 2

(3) for closed and open cell foams, respectively. In Eqs. 1, 2 and 3, AEc and Aσc are coefficients to be fit to experimental data. Figures 1 and 2 show the mechanical properties of Table 2 plotted against relative density and compare those data to the predictions of Eqs. 1, 2 and 3plotted for a range of values for the coefficients AEc and Aσc. The figures demonstrate that no single choice of the coefficients AEc and Aσc can yield accurate predictions of the material properties of steel foams across all manufacturing processes, and that even within a manufacturing process, the mechanical properties appear to depend on more than the relative density. These comparisons demonstrate that substantial further research is required to develop models that can predict the mechanical properties of steel foam produced by any of the currently available processes, and that any such models must incorporate mechanics of deformation beyond a characterization of the material relative density. 4 Example application of steel foam: Kremer et al. [8] reported success in mitigating local buckling of a thinwalled steel tube in bending by filling the maximum moment section of the tube with a moderate density (ρ = 0.40) steel foam. The composite tube exhibited higher peak load and more ductility than the empty tube. Because local buckling of structural members is essentially a manifestation of plate buckling, we have investigated the crushing and elastic critical loads of a steel foam plate simply supported on four edges and

E*FC628G51/)*+,-./0*1+2+./;HBIHBJ.D

)*+,-./0*1+2+./3./45267835/95,.87: ;8?.*,/@/A.=?:/B2*.51!B522/C*152D /

!"$ 9676 >@A/C6L >@A/#I$ >@A/C8,

!"#'

AH K#"! B

A K!"'' H B

!"#

!"!'

A K!"# H B

!/ !

!"#

!"$ !"% !"& 45267835/95,.87:

!"'

!"(

Figure 1: Comparison of predictive equations for cellular material elastic modulus to experimental measurements of elastic modulus for steel foams produced by a variety of processing methods. Experimental measurements for hollow sphere, composite and powder metallurgy foams are shown, all of which have closed cell morphology. G

or σf

)

FK0



(

)

F

(

H+.,=51I/63*+,-./0012/341/56378./9:8;3@! J!

maintaining sufficient stiffness in structural applications of steel foam will be a critical and challenging objective. Table 2 clarifies the nearly complete lack of material property characterization beyond compressive properties, and provides strong motivation for our efforts at cyclic, tensile, and shear measurements. Ashby [3] and Gibson & Ashby [6] have provided mechanics-based expressions that purport to predict the mechanical properties of cellular materials as a function of the relative density and base material properties. For example, the elastic modulus is related to the relative density by Ef = AEc 0.5ρ 2 + 0.3ρ (1) Es for closed cell foams, and the yield stress by σf = Aσc 0.5ρ 2/ 3 + 0.3ρ (2) σs

*+,-./0012/341/56378./9:8;3203/9018? @A18;3B1C0+93D3E0;C?3+-/9!3=963F5+0/6!F/553,+6/50G 3 # *+,-+018/3M7 M7 N9O/F81+93P+56 QP Q./FR.0+. *5+0/63BDE3,=S *5+0/63BDE3#J$ *5+0/63BDE3,19 T-/93BDE3,=S T-/93BDE3#J$ T-/93BDE3,19

!") !"( !"& !"$ !3 !

E! L#"! F

E L!"'' ! F

E! L!"# F

!"#

!"$ !"% !"& /9018?

!"'

!"(

Figure 2: Comparison of predictive equations for cellular material yield stress to experimental measurements of compressive yield for steel foams produced by a variety of processing methods. loaded in pure compression. The main objective of this preliminary study is to determine how varying the relative density of the steel foam, while holding constant the weight per unit area of the plate, affects the elastic critical and crushing loads of the plate. In order to hold the weight per unit length of the cross section constant while reducing the relative density of the walls (i.e. replacing the solid steel walls with steel

Proceedings of 2011 NSF Engineering Research and Innovation Conference, Atlanta, Georgia

Grant #1000334

€ €

(





€ €

One can observe that lowering the relative density results in a transition from instability to crushing, and that the value of ρ at which this transition occurs depends on the plate width. Figure 4 shows Pcry when the initial thickness of the plate is allowed to vary. There is again a transition between regimes in which the onset of elastic stability and crushing initiate a departure from linear response, and the curve delineating this transition depends on the relative density. The value of the Pcry is given by

 7986t s3  Pcrf = Pcry =  ρ  P = 31742t ρ1/ 2 s  yf

t s < 1.99ρ 3/ 4 t s ≥ 1.99ρ

(12)

3/ 4

)

By substituting the thickness and elastic modulus of the steel foam plate (Eqs. 4 and 5) into Eq. 7 the elastic critical load becomes 4 π 2 Es t 3s (8) Pcrf = 12 1− ν 2 bρ

(

higher buckling capacity and lower crushing capacity than a solid steel plate with the same total mass, geometry, and boundary conditions. Determining the actual capacity of the plate requires analysis of the postbuckling and inelastic buckling of the plate, both topics of current investigation by the authors. In order to further investigate the relationship between plate properties and the crushing and elastic critical loads, we allow the plate width and initial thickness to vary and define a response surface Pcry = min(Pcrf,Pyf), noting that this does not represent the capacity of the plate since it does not consider postbuckling or inelastic buckling behavior. Figure 3 shows this response surface for the case of variable plate width given by  3788838 b < 80 ρ −3/ 4  Pcrf = bρ . (11) Pcry =   Pyf = 595bρ1/ 2 b ≥ 1.99 ρ −3/ 4 

)

which depends inversely on the relative density, and therefore increases when the relative density is decreased since ρ < 1 for a foam. The crushing load of the plate is Pyf = Fyf t f b (9) which, by substitution of Eqs. 4 and 5, becomes Pyf = Fys t sbρ1/ 2 (10) which depends on the square root of the relative density and decreases when the relative density is decreased. Thus, a steel foam plate would be expected to have a

100

Crushing Pcry (kN)



foam walls), the thickness of the cross section has to be increased. The weight per unit area of a solid steel and steel foam plate are t s ρ s and t f ρ f respectively, where ts and ts are the solid steel and steel foam plate thicknesses, and ρs and ρf are the solid steel and steel foam densities. The constraint on the weight per unit € length of the member€ can then be expressed as, t (4) tf = s ρ which relates the thickness of the steel foam plate to the thickness of the solid steel plate and the relative density of the foam. The material properties of the steel foam are different from those of solid steel and depend on the base metal properties and the relative density. Gibson and Ashby [6] developed the expressions E f = Es ρ 2 (5) and Fyf = Fys ρ 3/ 2 (6) that relate the elastic modulus and yield stress of € the foam to the relative density. Although we showed in section 3 that equations such as these cannot predict the mechanical properties of steel foams across a wide variety of manufacturing processes, what will be shown in the following calculations is that the exponent in the expressions determines the interaction between steel foam relative density and the crushing and elastic critical load of the plate. The plate we investigate has b = 92mm and initial thickness ts = 1.73mm and the elastic modulus and yield stress of the base steel are assumed to be 200 GPa and 345 MPa respectively. The material is assumed to be perfectly plastic after yield. The elastic critical load of the plate is 4 π 2 E f t 3f Pcrf = € (7) 12 1− ν 2 b

Buckling

50

300 250

0

200

0.2

150

0.4 100

0.6 0.8 1

50

Plate Width (mm)

Relative Density

Figure 3: Crushing and elastic critical response surface for a simply supported plate loaded in pure compression and made of steel foam with varying relative density and width. Zones in which the response is governed by elastic instability and crushing are indicated.

Proceedings of 2011 NSF Engineering Research and Innovation Conference, Atlanta, Georgia

Grant #1000334

Pcry (kN)

100

Crushing

50 2.5 2

0

Buckling

0.2

1.5 0.4 1

0.6 0.8 1

Plate Thickness (mm)

Relative Density

Figure 4: Crushing and elastic critical response surface for a simply supported plate loaded in pure compression and made of steel foam with varying relative density and initial thickness. Zones in which the response is governed by elastic instability and crushing are indicated. 5 Conclusions: A review of the published literature on processing and material properties of steel foams shows that a variety of methods for fabricating steel foams have been developed and implemented at the laboratory scale over the past 10-15 years. Of these methods, powder metallurgy, sintering of hollow spheres, and composite powder metallurgy – hollow sphere processes appear to be the most promising for yielding materials of potential use in civil structural applications because of their relatively simple manufacturing processes, simple cell morphology, and potential to deliver high relative densities. Steel foams with a very wide range of material properties have been demonstrated, and these properties do not appear to correlate well with predictive equations that treat only the relative density as the material parameter. Finally, we have evaluated the loads at which a simply supported plate loaded in compression will crush and suffer elastic instability when made of steel foam, and have evaluated the dependence of those loads on the plate thickness and width, and relative density of the steel foam. We have found that a steel foam plate will tend to experience elastic instability at a higher load and crushing at a lower load than a solid steel plate of equivalent mass and subject to the same loading. 6 Acknowledgements: This work was supported by the United States National Science Foundation through grant #1000334. 7. References: [1] Adler J, Standke G, and Stephani G (2004). “Sintered open-celled metal foams made by replication

method - manufacturing and properties on example of 316L stainless steel foams.” Proceedings of the Symposium on Cellular Metals and Polymers (CMaP). Deutsche Forschungsgemeinschaft (DFG), 12-14 October 2004, Fürth, Germany, p.89-92. [3] Angel S, Bleck W, and Scholz P-F (2004). “SlipReactionFoamSintering (SRFS) - process: production, parameters, characterisation.” Proceedings of the Symposium on Cellular Metals and Polymers (CMaP). Deutsche Forschungsgemeinschaft (DFG), 1214 October 2004, Fürth, Germany. [3] Ashby M, Evans A, Fleck N, Gibson L, Hutchinson J, Wadley H. (2000) Metal Foams: A Design Guide. Butterworth-Heinemann. [4] Brown JA, Vendra LJ, and Rabiei A (2010). “Bending properties of Al-steel and steel-steel composite metal foams.” Metallurgical and Materials Transactions A. Online:1 July 2010. [5] Friedl O, Motz C, Peterlik H, Puchegger S, Reger N, and Pippan R (2007). “Experimental investigation of mechanical properties of metallic hollow sphere structures.” Metallurgical and Materials Transactions B. 39(1):135-146. [6] Gibson L, Ashby M. (1999) Cellular solids: Structure and properties-second edition. Cambridge University Press. [7] Gong L, Kyriakides S, Jang W-Y. (2005). “Compressive response of open-cell foams. Part I. Morphology and elastic properties.” International Journal of Solids & Structures 42, 1355–1379. [8] Kremer K, Liszkiewicz A, Adkins J. (2004) “Development of steel foam materials and structures.” Tech. rep., Fraunhofer USA Delaware Center for Manufacturing and Advanced Materials, 9 Innovation Way Newark, DE 19711, US. [9] Muriel J, Sanchez Roa A, Barona Mercado W, and Sanchez Sthepa H (2009). “Steel and gray iron foam by powder metallurgical synthesis.” Suplemento de la Revista Latinoamericana de Metalurgia y Materiales. 2009. S1(4):1435-1440. [10] Neville BP and Rabiei A (2008). “Composite metal foams processed through powder metallurgy.” Materials and Design 29:388-396. [11] Park C and Nutt SR (2000). “PM synthesis and properties of steel foams.” Materials Science and Engineering A. A288:111-118. [12] Park C and Nutt SR (2001). “Anisotropy and strain localization in steel foam.” Materials Science and Engineering A. A299:68-74. [13] Park C and Nutt SR (2002). “Strain rate sensitivity and defects in steel foam.” Materials Science and Engineering A. A323:358-366. [14] Tuchinsky L (2005). “Novel fabrication technology for metal foams.” Journal of Advanced Materials. 37(3):60-65.

Proceedings of 2011 NSF Engineering Research and Innovation Conference, Atlanta, Georgia

Grant #1000334

[15] Rabiei A and Vendra L J (2009). “A comparison of composite metal foam's properties and other comparable metal foams.” Materials Letters 63:533536. [16] Verdooren A, Chan HM, Grenestedt JL, Harmer MP, and Caram HS (2005). “Fabrication of low density ferrous metallic foams by reduction of ceramic foam precursors.” Journal of the Materials Science. 40:43334339. [17] Verdooren A, Chan HM, Grenestedt JL, Harmer MP, and Caram HS (2005). “Fabrication of low density ferrous metallic foams by reduction of chemically bonded ceramic foams.” Journal of the American Ceramic Society. 89(10):3101-3106. [18] Weise J, Beltrame Derner Silva G, and Salk N (2010). “Production and properties of syntactic steel and iron foams with micro glass bubbles.” In Proceedings of MetFoam 2009, Bratislava, Slovakia.

Proceedings of 2011 NSF Engineering Research and Innovation Conference, Atlanta, Georgia

Grant #1000334

Suggest Documents