D.H. Lammlein PhD Dissertation - Vanderbilt University

More documents

Recommendations

Info

For full circumferential welds, the eccentricity of the experimental setup, shown in Figure 104, is approximately 0.02”. This level of eccentricity results in significant and unacceptable changes in the vertical position of the sphere surface relative to the tool, and thus unacceptable variations in weld depth. An in-process, force feedback control algorithm was therefore employed for the conical tool experiments as a means of maintaining a constant penetration depth around the entire sphere. The vertical position was adjusted based on torque signals received from a rotating cutting force dynamometer oriented in series with the tool. For the conical type tools it was found that desired position was better maintained if torque, rather than force, was used as the input in the decision making algorithm. An auto-zeroing procedure is performed over the start of the weld to determine the height of the material surface here relative to the tool. The tool is then plunged to the assigned penetration depth and the weld traverse is started. After 1” of weld traverse the measured torque from the dynamometer is set as the desired torque and from this point forward this torque value is maintained by vertical adjustments of tool position using the weld machine's vertical axis motor. Figure 123 in a later section shows the weld torque increasing with penetration depth at the 1000rpm, 10ipm parameter setting. The relationship between weld depth and process forces will be discussed later. Figure 105 and Figure 106 below demonstrate the torque control process for two welds runs(1000rpm, 10.4ipm, 0.095” weld depth). In the upper frame the axial force (blue) and desired axial force (green) are shown. The desired axial force is not used in the control algorithm and is shown for comparison with the measured axial force and the torque panel below. Torque was found to be a more reliable indicator of depth. These relationships are discussed later. The middle panels in Figure 105 and Figure 106 show the measured torque (blue) and desired torque (green). It can be seen in the figures that the desired torque, set after 1” of weld traverse, is followed closely. The bottom panel of the two figures is the vertical position of the tool relative to the height of the material surface at the weld plunge site (weld start). A linear encoder is used to track movements relative to this initial calibration. In this manner the vertical adjustments made by the tool to maintain the desired torque can be seen. As torque is highly sensitive to plunge depth and a constant torque is closely maintained throughout the weld traverse, it can be concluded that tool penetration depth is closely maintained throughout the weld traverse. 118
The adjustments in vertical position in the bottom panel of the figures therefore shows the eccentricity of the setup or how the material surface height at the tool engagement site changes relative to its initial measurement as the weld piece is rotated. Additional details and description of the torque control method used in these experiments can be found in Longhurst et al. [26]. Figure 105: Dynamometer data for a butted configuration, sphere weld (1000rpm, 10.4ipm, 0.095” depth) using a 100º conical tool (snub-cup). The top frame shows measured force(blue) and desired force (green). The desired force is presented for comparison and not used in the control algorithm. The middle panel shows measured torque(blue) and desired torque (green). Toque is highly sensitive to plunge depth. The desired depth is followed closely by maintaining weld torque via vertical adjustments of tool position. The weld is a full circumferential weld and the eccentricity of the setup can be seen in the lower frame which shows the vertical position of the tool relative to its initial calibration on the material surface. 119
Page 1 and 2:
FRICTION STIR WELDING OF SPHERES, C
Page 3 and 4:
TABLE OF CONTENTS TABLE OF FIGURES
Page 5 and 6:
Appendix ..........................
Page 7 and 8:
Figure 16: Left(FE streamline), Mid
Page 9 and 10:
Figure 51: A thermal camera image t
Page 11 and 12:
for a similar conventional weld (3/
Page 13 and 14:
Figure 102: Tapered retraction proc
Page 15 and 16:
setting are indicated. Note that st
Page 17 and 18:
Unsupported welds of this kind with
Page 19 and 20:
Figure 157: A thermal camera image
Page 21 and 22:
ottom sample is a tungsten inert ga
Page 23 and 24:
N - newton (force unit) n - visco-p
Page 25 and 26:
INTRODUCTION Friction Stir Welding,
Page 27 and 28:
educed conduction lengths. In small
Page 29 and 30:
CHAPTER I LITERATURE REVIEW: COMPUT
Page 31 and 32:
frictional condition at the interfa
Page 33 and 34:
contact. The contact area can inclu
Page 35 and 36:
For horizontal sections, a heat inp
Page 37 and 38:
(1.11) where r p is the probe (pin)
Page 39 and 40:
This efficiency term, η pd , is kn
Page 41 and 42:
The heat input on the tool surface
Page 43 and 44:
The added complication of these met
Page 45 and 46:
Figure 5: Lateral cross sections of
Page 47 and 48:
Flow stress in aluminum alloys is d
Page 49 and 50:
Figure 8: Three flow fields a) rota
Page 51 and 52:
and post-weld cool-down. This model
Page 53 and 54:
Williams et al. [60] use a 2D-axisy
Page 55 and 56:
Figure 16: Left(FE streamline), Mid
Page 57 and 58:
Figure 18: Temperature contour over
Page 59 and 60:
Figure 22: Temperature distribution
Page 61 and 62:
Figure 24: Lateral velocity contour
Page 63 and 64:
Figure 26: Experimental(white) vers
Page 65 and 66:
Figure 28: Top view schematic of tu
Page 67 and 68:
[5] Taylor RE, Groot H, Goerz T, Fe
Page 69 and 70:
[41] Heinz B., Skrotzki B., Eggler
Page 71 and 72:
CHAPTER II COMPUTATIONAL ANALYSIS O
Page 73 and 74:
Figure 30: Blind t-joint setup with
Page 75 and 76:
Figure 33: Process forces and later
Page 77 and 78:
Figure 35: Experimental Welds, Axia
Page 79 and 80:
Figure 37: FEA boundary conditions
Page 81 and 82:
A 4500 N axial force is applied eve
Page 83 and 84:
axial force with lateral offset dur
Page 85 and 86:
Figure 42: Contour plot of von Mise
Page 87 and 88:
Figure 44: Deformation contour for
Page 89 and 90:
Figure 46: Quarter-plate model for
Page 91 and 92: Figure 49: Contour plot of deflecti
Page 93 and 94: where P is the weld power (W), Q is
Page 95 and 96: Figure 52: Fluent CFD model zones.
Page 97 and 98: Figure 55: Thermal contour of weld
Page 99 and 100: Figure 59: Thermal contour of weld
Page 101 and 102: Figure 61: Velocity vectors for 0.0
Page 103 and 104: Figure 65: Contour of velocity magn
Page 105 and 106: CHAPTER III THE APPLICATION OF SHOU
Page 107 and 108: Figure 66: A shoulder-less, conical
Page 109 and 110: Figure 68: Typical axial(Z-axis) fo
Page 111 and 112: Figure 71: 90° conical tool macros
Page 113 and 114: Figure 75: Run 3, 80° conical tool
Page 115 and 116: Figure 79: (90º tool) Ultimate ten
Page 117 and 118: Figure 83: (90º tool) Lateral forc
Page 119 and 120: Figure 87: (80º tool) Lateral forc
Page 121 and 122: The Eulerian, finite volume, CFD so
Page 123 and 124: Figure 92 shows the increasing elem
Page 125 and 126: Figure 91: CFD model geometry consi
Page 127 and 128: Figure 95: Lateral cross-section of
Page 129 and 130: Figure 96: Attempts a probe tapered
Page 131 and 132: CHAPTER IV THE FRICTION STIR WELDIN
Page 133 and 134: applications for its high strength
Page 135 and 136: Figure 97: Experimental weld sample
Page 137 and 138: Figure 99: Split protrusion type de
Page 139 and 140: with a cupped recess. It is however
Page 141: of tapered retraction (i.e. move th
Page 145 and 146: manners the assigned depth was main
Page 147 and 148: Figure 108: Lateral macrosections f
Page 149 and 150: Figure 110: (Supported, cupped tool
Page 155 and 156: apparent strength of 26% parent whi
Page 157 and 158: with a 100º conical tool (0.025”
Page 159 and 160: Figure 119: Macrosection view of an
Page 161 and 162: and tooling. Torque is highly depen
Page 163 and 164: locally in proportion to the local
Page 165 and 166: Figure 126: (threaded tool) Contour
Page 167 and 168: Figure 128: (conical tool) Geometry
Page 169 and 170: Figure 131: (threaded tool) Lateral
Page 171 and 172: The results presented here show tha
Page 173 and 174: Figure 136:(from left: Tapered retr
Page 175 and 176: [8] Metallurgical analysis of ablat
Page 177 and 178: CHAPTER V FRICTION STIR WELDING OF
Page 179 and 180: Collaboration between Advanced Meta
Page 181 and 182: eccentricity, the method of interio
Page 183 and 184: gas tungsten arc welding of small d
Page 185 and 186: Full penetration welds of 4.2” (1
Page 187 and 188: Figure 143: The curvature of the pi
Page 189 and 190: Figure 144: Experimental axial forc
Page 191 and 192: Together, a fine wall thickness tol
Page 193 and 194:
Figure 147: Axial force history for
Page 195 and 196:
The weld macrosections show complet
Page 197 and 198:
Figure 154: Macrosections of welds
Page 199 and 200:
Figure 157: A thermal camera image
Page 201 and 202:
Figure 159: Average tool shank temp
Page 203 and 204:
471,146 faces. The meshes are fine
Page 205 and 206:
Figure 162: A closeup view of mesh
Page 207 and 208:
4 2 3 4 5 16 " 1.745 2 0.24
Page 209 and 210:
applied heat locally in proportion
Page 211 and 212:
Figure 164: Modeled temperature con
Page 213 and 214:
Page 215 and 216:
Page 217 and 218:
Figure 172: Model pathlines for the
Page 219 and 220:
Figure 174: Model pathlines for the
Page 221 and 222:
later speed setting, the tool welds
Page 223 and 224:
Figure 178: Zoom view of a macro ta
Page 225 and 226:
Figure 180: The superficial appeara
Page 227 and 228:
Figure 182: The rotary welding appa
Page 229 and 230:
Figure 185: A diagram illustrating
Page 231 and 232:
CONCLUSION AND FUTURE WORK Conclusi
Page 233 and 234:
new FSW process environment is reso
show all

D.H. Lammlein PhD Dissertation - Vanderbilt University

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?