The lithosphere east of the Gorda ridge, often called the Gorda plate but referred to here as the Gorda deformation zone (Gorda DZ), is constricted as it spreads eastward to be subducted beneath the North American plate. For the plane strain conditions defined by the seismicity, the distortion of the Gorda DZ is analogous to the ironing of metal cups. Here the ''punch'' boundary is the Blanco fracture zone and the ''die'' that does the cutting is the Mendocino transform fault from the Gorda ridge to the Mendocino triple junction. This analogy is useful in analyzing this problem but is not identical due to differences in boundary constraints. The deformation of the Gorda DZ is determined for an elastoplastic lithosphere with a von Mises yield criterion. Plate motion is constrained to be parallel both to the Blanco fracture zone and to the Mendocino transform fault west of 125 ¿W. The Gorda ridge is considered a free boundary (unlike the ironing analogy), and the lithosphere is pulled through the constriction from the trench. Any nonlinear feedback between distortion and ridge-push forces is not considered. Traction due to slip along the Mendocino transform is varied to fit both seismicity and plate distortion. For these conditions, the amount of plate movement required before a constant resistance is achieved depends upon the ratio of elastic moduli to plastic yield stress and may vary from hundreds of meters to kilometers. The minimum slab pull for steady movement occurs when the transform is weak (''lubricated'') and is 60% of the average plate yield stress.

The yield stress is determined using the temperature structure inferred from plate age and the laboratory constraints on brittle and ductile strength for olivine. For an average plate age of 3 m.y. the yield stress is about 550 MPa, so the pull required is 330 MPa. When the transform is about as strong as the lithosphere, the pull required is twice this value. The stress directions in the Gorda DZ can be inferred from both deformation and seismicity. The current seismic data are not precise enough to differentiate between a strong and weak transform, but the deformation data are sufficient. The constraints placed by distortion of magnetic isochrons on the modeling results indicate a mixed history. About 5 Ma, changes in the position of the pole of rotation of the Juan de Fuca system began to increase the convergence of the Gorda DZ with the Mendocino transform fault, increasing the effective normal stress and brittle shear stress across the fault. A strong fault is evident in the deformation data until about 2 Ma, and this corresponds to the time when convergence stabilized. With a stable configuration the analysis indicates that about a kilometer of relative slip reduced the apparent strength of the transform to negligible values compared to the strength of the plate. ¿ American Geophysical Union 1992