# CIAM Fig. 6 digitization: q and fuel flow only

Status: approximate candidate values read from a plotted raster; not source telemetry and not exact measurements.

Source: R. T. Voland et al., “CIAM/NASA Mach 6.5 Scramjet Flight and Ground Test,” AIAA-99-4848 (1999), Fig. 6, physical PDF page 4, NASA NTRS record 20040087160. Source PDF SHA-256: `8a9c844a03d20ee4b734e855cfe177f12c792f541b248293abb7096204837781`. Figure 6 crop: `analyses/rendered-S-PROP-004-NASA-CIAM-MACH65-2026-09-29/figure-6-4x.png` (920 × 588 px; 4× source crop). The original PDF page was inspected at full page for axis labels, trace identity, and context; the candidate pixel readings were taken from the 4× crop.

## Calibration

The graph plotting rectangle in the crop is calibrated from the visible axes: x = 38 px at 0 s and x = 753 px at 120 s; y = 485 px at 0 and y = 15 px at the upper plot boundary. Thus `time_s = (pixel_x − 38) × 120 / 715`. Dynamic pressure uses the left axis, 0–7 bar: `q_bar = (485 − pixel_y) × 7 / 470`. Fuel flow uses the right axis, 0–0.18 kg/s: `fuel_kg_s = (485 − pixel_y) × 0.18 / 470`. Stored pixel coordinates are approximate trace-marker centers; all values are calculated from those coordinates and rounded to 0.01 bar or 0.001 kg/s in the CSV.

## Method and precision

I visually followed only the triangle-marked dynamic-pressure trace and cross-marked fuel-flow trace, recording each discernible plotted marker/vertex, then converted its center using the linear axis transforms above. This is a manual raster digitization, not a calibrated instrument-data extraction. Coordinate placement is uncertain by roughly ±1–2 crop pixels (about ±0.17–0.34 s, ±0.015–0.030 bar, or ±0.0004–0.0008 kg/s before line thickness, antialiasing, and ambiguity are considered). Reported rounded precision is for convenient plotting only and must not be interpreted as measurement accuracy.

The source has no raw time-series table, error bars, channel calibration uncertainty, or repeat-flight statistics. Do not imply statistical confidence or exact telemetry. A future redraw should identify these as “approximate visual readings from Fig. 6”.

## Ambiguities and discontinuities

- Dynamic pressure rises sharply from the origin to about 1.9 bar by roughly 6.5 s; the origin and first trace point are close to the plot axes, so the earliest slope is especially approximate. A shallow minimum around 35–42 s is near 0.8 bar.
- Fuel flow lies on/near zero through roughly 39 s, rises sharply to about 0.16 kg/s near 42 s, then changes rapidly around 46–48 s. Individual vertices there are close together and may be displaced by the crossing/overlap with other plotted curves, so their time/value readings should be treated as coarse.
- The fuel trace declines to the baseline around 88 s and steps back to approximately 0.05 kg/s around 94–95 s. These abrupt transitions are present in the plotted line; the reason and exact switching times are not inferred here. Baseline readings indicate visually unresolved near-zero values, not proof of mathematically exact zero flow.
- Short straight segments between plotted points are only the source plot’s connecting lines. Do not interpolate them as recovered high-frequency data or infer unshown physical dynamics.

This artifact digitizes two plotted channels only. It does not infer a complete engine/system design, energy recovery, or efficiency. The source’s primary audit separately notes this was one flight, that fueling lasted about 77 s, and that the plot contains no pointwise source data or uncertainty bars. Cite the source and NASA NTRS record when using a derivative plot. Figure-artwork rights are not independently cleared; this is a new approximate redraw dataset, not permission to republish the original figure.

The paper text on p.4 says fuel injection began at about 38 s and that fuel flow continued to the 115 s flight-termination event, apart from a short interruption around 90 s. The raster trace's first visible rise reads about 39–42 s under this calibration; preserve that mismatch as plot-reading uncertainty rather than forcing the trace to the prose value. Candidate points in the CSV have been checked against the listed pixel transforms by `validate_digitization.py`.