Raymond J. Emrich

Investigation of the Shock‐Tube Boundary Layer by a Tracer Method

A dark‐field oscillatory spark microphotography technique has been developed to study with tracers the structure of the gas boundary layer in the shock tube. By using these particles as tracers, velocity profiles of the boundary layer both in laminar and turbulent regions are measured and the velocity fluctuations observed. The time and space resolution of this technique is high. Laminar velocity profiles show a significant deviation from theoretical expectations. The turbulent average velocity profiles have the usual 1/7th power variation with distance from the wall in the top 90% of the layer. The velocity of the flow within a few microns from the wall is sometimes much greater than the value predicted by Mirels assuming zero velocity at the wall. A possibility of slip flow in the shock tube is discussed. This technique allows measurements to be made as close to the wall as the size of the particles diffraction pattern, which is about 1 μ.

Wall Effects in Shock Tube Flow

In order to study wall effects in shock tube flows, measurements have been made of shock strength, of average density across the tube, and of pressure at the walls. These measurements are compared with explicit predictions of Trimpi and Cohen and of Mirels and Braun. Under conditions where the theories should apply, they predict shock attenuation and values of the flow variables sufficiently well to justify as essentially correct the assumptions that the turbulent boundary layer produced by the shock is similar to the boundary layer observed in steady flows and that waves generated by wall effects can be considered as one‐dimensional. A need for some modification of the theories is indicated, however, by lack of quantitative agreement in certain cases. Considerable mixing of the hot and cold gases is observed to occur. Neither theory attempts to explain the mixing nor takes into account its effects on wave propagation.

Turbulent Spots and Wall Roughness Effects in Shock Tube Boundary Layer Transition

Turbulent spots have been identified as a transition mechanism in shock tube boundary layer flows for a range of weak shocks at initial pressures near one atmosphere in air, using thin‐film wall temperature gauges. The shape and growth rate of the spots is consistent with that found by other investigators in sub‐ and supersonic steady flows. A study of naturally occurring and artificially generated turbulent spots, and of flow tripping by two‐ and three‐dimensional roughness elements, has shown the existence of an unconditionally stable region behind the shock, in which finite disturbances will not cause flow breakdown or transition to turbulence. The limit of this stable region is given by Reδ = 1.7 ± 0.3 × 103, based on boundary layer thickness at 99% of free‐stream velocity. In the absence of artifical perturbations, laminar flow is seen to persist for times as much as a factor 5 longer than those seen by previous investigators. These maximum transition times were limited by flow tripping at wall disco...

Flow field measurement by tracer photography

Methods of inserting and photographing tracers in an air stream are described. Successive positions of micrometer-sized oil drops are determined in centimeter-sized fields of a low power microscope by triple flash illumination at known intervals of 10 to 500 microseconds. Tracer positions read from the picture are entered into a computer and velocities calculated throughout the field. Some examples of computer displays of velocity fields are presented.

Tracer Study of Pipe Flow Behavior to within Two Microns of the Wall

A high precision, high spatial resolution tracer technique was used to study low noise, steady air flow in a 1½ by ½ in. rectangular pipe with particular attention to the flow several microns from the pipe floor. The flows investigated had velocities which correspond to pipe Reynolds numbers of 16–4300. Experimental velocity profiles for the flows extrapolate to the no‐slip condition at the floor. However, close attention to the measurements of the flow 2–10μ from the floor reveals a preponderance of data points lying on the high velocity side of the profile. The data are believed to mirror a real deviation of the flow velocity from the theoretical profile in the region of the floor and greater than 2 μ from it. The cigarette smoke tracers could be located within a micron from the floor, and the measuring error in determining the velocity varied from 0.02 mm/sec in flow of 5 mm/sec and less, to 0.15 mm/sec in a flow of 50 mm/sec.

Schlieren and interferometric study of a laser triggered air spark in the nanosecond regime

The growth of the plasma column and the electron density in an atmospheric air spark are studied by photographs. A 1‐ns nitrogen laser provides the light source and the trigger. The relative timing is controlled by an adjustable light path length. It was found that the spark bridges the 0.5‐mm gap during the first 5 ns. The plasma column diameter then grew linearly with time for 25 ns at a speed of 10.5 km s−1. The electron density in the ionized channel is found from interferograms; it reached a maximum value of 6.5×1019 cm−3 at 8 ns.

Precision Measurement of Parabolic Profile for Laminar Flow of Air between Parallel Plates

Photography of tiny oil drop tracers carried by air flowing in a rectangular pipe has been used to measure the fluid velocity at various distances from the wall. Measuring accuracy of better than±0.5% over most of the profile yields agreement with the predicted parabolic profile. The measurements are carried out in an all‐glass section of pipe of cross section 9.5 × 102 mm; the pipe flow Reynolds number is 15, and the duration of the measurement is 0.2 sec.

Observation of Shock Formation and Growth

Transient one‐dimensional flows have been produced by accelerating a piston in a gas‐filled tube. Accelerations of the order of 2.5×104 m/sec2 produced shocks within 5 m of the initial piston position. The piston was propelled pneumatically. Shocks formed at the head and in the interior of the compression waves. By optically detecting the piston and shock positions and recording the corresponding times on a rotating drum chronograph with microsecond accuracy, the properties of the compression wave and resulting shock were determined. The formation and growth of shocks were calculated from the compression wave measurements using the method of characteristics and the assumptions employed by Chandrasekhar and Friedrichs. The observed and calculated shock properties were in reasonably good numerical agreement. Consistent lagging of the observed behind the calculated shock positions (order of 0.05 m at 5 m from initial piston position) is attributed to the interaction of the flow with the tube walls.