**3. Results**

For this project, one of the flexures as a key component of a prototype microspline of an asteroid gripping device under development at NASA/JPL for the Asteroid Redirect Mission [5,6]. The delicate geometry of the selected flexure, as shown in Figure 1, presented considerable challenge to most machine tools. There is an option to (1) machine or fabricate the flexure with the strengthening tabs attached (Figure 1a) and then remove the tabs to finish the part or (2) machine the flexure without the tabs (Figure 1b).

#### *3.1. Waterjet Cutting*

The prototype flexure made of 6061 T6 aluminum using the MicroMAX was originally cut to demonstrate its performance versus that of the wire EDM. The EDM process conducted at JPL was carried out in three passes in order to minimize the damage resulted from the induced heat-affected zone (HAZ). The comparison showed that the cost ratio for machining the part was 14:1 in favor of the waterjet, leading to a 93% cost reduction [3].

Figure 2a shows the micrograph of the aluminum flexure (0.81 mm THK) cut on the MicroMAX using the 7/15 nozzle with the Barton 240 or 220UT mesh garnet with a mean particle size of 60 μm. The pump pressure was 380 MPa and the abrasive mass flow rate was 73 g/min. The cutting time was 2.3 min. The geometry of the flexure element including the horizontal and semi-circle segments were inspected under the microscope. From the micrographs, the following features of the flexure are inspected:

	- Smoothness and the presence of chipping
	- Edge taper

Figure 2b shows the superimposition of the tool path onto the flexure element. The overall match between the tool path and the flexure element is excellent, indicating that there is no distortion of the flexure element in terms of bending and/or rotating in the X-Y plane.

Magnified views of the two small areas in the middle span and the right end loops of Figure 2a, were selected to compare in detail the flexure element and the tool path, as shown in Figure 2c,d. The areas were chosen because they were farthest away from the two anchoring points and least supported. The selection was made to show the worst mismatches in those two segments of the entire element. Yet, the match shown in Figure 2c is excellent. A very slight mismatch is observed in Figure 2d. The maximum mismatch was measured to be about 0.1 mm, part of which is attributed to the error in overlaying of the tool path onto the micrograph. Figure 2c,d show nearly no macro distortion induced by the waterjet cutting process, clearly demonstrating the advantages of cold cutting and low-side-force exertion. In fact, previous investigation verified that waterjet, as opposed to wire EDM, preserved the structural and chemical integrity of parent materials [8].

**Figure 2.** Aluminum flexure—MicroMAX.

The same flexure was then cut from a stainless-steel sheet (0.76 mm THK). The same cutting parameters for cutting the aluminum flexure were used. The cutting time increased to 3.4 min. Figure 3 illustrates the micrograph of the stainless-steel flexure. The flexure element displays no heat- and/or mechanically-induced distortion. Using the 7/15 nozzle, a 2/3-size but not a half-size flexure was successfully cut as the kerf width of the 7/15 nozzle is larger than the gap between elements of the half-size flexure.

**Figure 3.** Stainless steel flexure—MicroMAX.

Full-size flexures were cut on CBA's OMAX 5555. The same setup for the 7/15 nozzle was used. Figure 4 shows micrographs of two flexures cut on aluminum (0.81 mm THK) and stainless steel (0.64 mm THK). In Figure 4a,b the overall geometry shows no noticeable distortion of the flexure element. Minor mismatches between the tool path and the flexure element are however observed in the zoomed-in micrographs of the aluminum flexure as shown in Figure 4c,d that correspond to the mid-span and right end segments of the second and third loops from the top. In the mid-span (Figure 4c), an undercut section about 9 mm long was observed on one of the horizontal segments (third row down. The maximum undercut is about 0.15 mm. Since this is the only occurrence throughout the entire flexure element, it is most likely just an outliner. The maximum mismatch at the two ends is about 0.1 mm. As shown in Figure 4d, the Y-positions of the first loop are slightly below its designed positions marked by the tool path. In other words, this corresponds to the small rotational (clockwise) distortion about the axis of the X-Y plane. Note that no such rotational distortion was observed on the part cut with the MicroMAX (Figure 2d). Comparison of Figure 4c,d with Figure 2c,d indicated that the cutting accuracy is slightly better for the MicroMAX than for the OMAX 5555. This is expected as MicroMAX was specifically designed and constructed for meso-micro machining.

**Figure 4.** Metal flexures—5555 JMC (CBA).
