Benjamin Y. H. Liu

Experimental observation of aerosol deposition in turbulent flow

The deposition rate of aerosol particles in turbulent flow has been measured in vertical pipes for nominal Reynolds numbers of 10,000 and 50,000 and for particle dia. of 1·4–21 μm using monodisperse aerosols generated by the vibrating orifice monodisperse aerosol generator. The experiments covered a τ+ range of 0·21–774, where τ+ is the dimensionless particle relaxation time. For small τ+ values, the deposition velocity was found to increase rapidly with increasing particle size, reaching a peak at τ+ = 30. Beyond τ+ = 30, the dimensionless deposition velocity, V+, showed only a moderate dependence on τ+, decreasing slightly at higher τ+ values. Below τ+ = 10. V+ was found to increase as the second power of particle relaxation time and as the fourth power of particle size. The experimental results are in good agreement with several of the existing theories in the region below τ+ = 30. However, the theories are inadequate to explain the results beyond τ+ = 30.

A submicron aerosol standard and the primary, absolute calibration of the condensation nuclei counter

Abstract A submicron aerosol generator has been described. The generator combines pneumatic atomization with electrostatic classification to produce uniform particles from 0.01 to 0.5 μm at concentrations up to 10 6 particles/cc. The particle diameter accuracy of the aerosols is about 2%, and the concentration accuracy about 5%. This submicron aerosol generator has been used to produce NaCl aerosols as absolute calibration standards for condensation nuclei counters. Calibration results are presented showing that the Environment/One (Model Rich 100) counter has a linear response at concentration levels up to 130,000 particles/cc, but the indicated concentration based on the accepted Nolan-Pollak standard is lower than the acutal concentration by a factor of 2.5. Above an aerosol concentration of 130,000 particles/cc, the instrument response becomes nonlinear, but the response is independent of particle size (0.011–0.15 μm in diameter) in both the linear and nonlinear ranges. Similar studies of the General Electric counter showed that the instrument response is linear to the highest concentration level (65,000) used in the studies. Results of these experimental findings are discussed, and their implications in coagulation-coefficient and particle-diffusion measurements are considered.

Theoretical Study of Aerosol Filtration by Fibrous Filters

A theoretical analysis of filtration mechanisms has been made for fibrous filters in the region of maximum penetration. The theory is based on a boundary layer approach using the Kuwabara flow field to account for the interference effects of neighboring fibers. An improved expression for the diffusion and interception filtration efficiencies has been derived that compares well with the existing theories. A comparison of the developed theory with experimental data also confirms the validity of the present work. *This paper is based on the thesis of K. W. Lee in partial fulfillment of the requirements for the Ph.D. degree at the University of Minnesota.

The aerosol size distribution of Los Angeles smog

A total of 342 atmospheric size distributions have been measured at the California Institute of Technology in Pasadena, CA during August and September 1969 over the size range 0.003–6.8 μm using the Minnesota Aerosol Analyzing System in collaboration with other investigators making chemical, meteorological, mass, and optical measurements. This paper presents characteristics of the aerosol size distributions and discusses the significance of these findings. It also presents a new computer-compatible system of nomenclature and a new system for separating and analyzing the size distribution in one decade ranges. In addition to the usual log ΔN/ΔDp versus log Dp plot, it was found that a plot of ΔV/Δ log Dp versus log Dp was very useful for characterizing the distribution of particles larger than 0. 05 μm. From the later volume distribution plot, it was found that the smog aerosol was universally bimodal with the saddle point in the 1- to 2-μm size range and with the volume fraction for the individual size distributions below 1 μm being characterized with a grand correlation coefficient of 0.9719 with log normal distributions. The grand log normal distribution had a geometric mean of 0.302 μm and a geometric standard deviation of 2.25. The grand average volume below 1.05 μm was 34 μm3/cm3 and the volume above 1.05 μm was 24.1 μm3/cm3. For sizes above 1 μm, the volume distribution increases steadily up to 6.8 μm, the largest size measured by the MAAS. Comparisons with Lundgren impactor data suggest that this larger mode peaks at 7–10 μm and then drops sharply at about 15 μm. The grand average number distribution log ΔN/ΔDp vs. log Dp was very similar to the grand average measured by Clark and Whitby in Minneapolis in 1966. The number distribution can be characterized reasonably well in the size range from 0. 2 to 7 μm by ΔN/ΔDp = 0. 4 VTDp−4, where VT = 58. 1 μm3/cm3, the total volume fraction. The volume fraction smaller than 1. 05 μm was found to correlate well with light scattering, solar radiation, and, to a lesser degree, with ozone. Analysis of the data and comparisons with laboratory studies of coagulating aerosols suggest strongly that, during the daytime when there is even light smog, most of the aerosol mass, in the size range smaller than 1 μm, is contributed by condensation of photochemical reaction products on particles larger than 0. 1 μm rather than from coagulation of smaller particles. There is considerable evidence that, while most of the mass of particles smaller than 1 μm is contributed by photochemical reactions during smog periods, most of the mass of particles larger than a few microns comes from other sources. While the aerosol mass of particles smaller than 1 μm peaks at about noon, the maximum in the condensation nuclei count peaked at about 1500 at values on the order of 1. 8 × 105/cm3. On the other hand, in Minneapolis, when there was no volume peak at noon, the condensation nuclei count peaked at 1. 8 × 106/cm3. The third paper in this series shows that, because the coagulation coefficient between 0. 01-μm particles and the 0.3-μm particles of the lower volume mode is much higher than for 0.01-μm particles with themselves, the presence of the relatively large volume below 1 μm in Los Angeles smog decreases the maximum condensation nuclei count observed compared to cities with less severe air pollution.

On the performance of the electrical aerosol analyzer

Abstract The electrical aerosol analyzer has been studied using monodisperse aerosols. The aerosols were generated by the electrostatic classification principle described by Liu and Pui (1974a) and by the photo-chemical conversion of gaseous SO2 into sulfuric acid droplets. The measuring range of the instrument was found to be from 0·006 to 1·0 μm in particle dia. and the sensitivity was found to vary from 0·01 to 400 pA/(106 particles/cm3) over the range of the instrument. Data are presented showing the performance of the device as a function of particle size and as a function of the operating conditions of the diffusion charger.

Experimental study of particle deposition in bends of circular cross section

The deposition efficiency of liquid particles in tube bends of circular cross section has been measured for flow Reynolds numbers of 100, 1000, 6000, and 10,000. The particle Reynolds number, Re p, was in the range 0.6–3.9 for the laminar flow cases (i.e., Re = 100 and 1000), whereas for the turbulent flow cases (i.e., Re = 6000 and 10,000) Re p was in the range 1.3–12.7. Bends constructed of stainless steel and glass tubes of different diameters were used. The experiments were performed using monodisperse aerosols generated by the vibrating orifice aerosol generator. The results were in good agreement with the theory of Cheng and Wang for Re = 1000, but differed from theory for Re = 100. For the turbulent cases, no dependence was found on the flow Reynolds number and an exponential curve of deposition efficiency versus Stokes number was fitted to the experimental results. A theoretical justification of the form of the curve is given in the paper.

Electrical neutralization of aerosols

The charge neutralization process for aerosol particles in a bipolar ionic atmosphere has been studied theoretically and experimentally. The degree of neutralization was found to depend on the parameter, Nt, where N is the ion concentration and t is the neutralization time. For particles in the 0·001–100, μm dia. range, the maximum Nt product needed for charge neutralization was found to be 6 × 106 and 1 × 105 (ions/ml) (sec) respectively for the two limiting cases studied, i.e. a ⪢ λanda < λ, where a is the particle radius and λ is the mean free path of the small ions. Experimental studies were made on four radioactive krypton-85 neutralizers with activities of 0·5,1, 2 and 10 mCi. The equilibrium ion concentration was found to range from 1·2 × 106to 8·8 × 106ions/ml. The performance of these devices as aerosol charge neutralizers was investigated using monodisperse DOP (di-octyl phthalate) aerosols of 2·53 μm dia. The results are in good agreement with the theoretical predictions.

An aerosol generator of high stability

This paper reports on the use of a syringe pump to provide a constant liquid flow to an atomizer to obtain an aerosol generator of high stability. This syringe-pump atomizer has also been used with a vaporizer-condenser to generate monodisperse DOP (di-octyl phthalate) aerosols in the 0.032 to 1 mum diameter range.

The aerosol mobility chromatograph: A new detector for sulfuric acid aerosols

Abstract A new instrument has been developed for measuring sulfuric acid aerosols. The instrument is called an Aerosol Mobility Chromatograph since it is based on the electrical mobility of aerosol particles and operates in a way similar to that of the conventional liquid or gas Chromatograph. The particle diameter range of the instrument is from 0.005 to 0.2 μm and the sensitivity (for detecting monodisperse sulfuric acid aerosols), from 0.01 to 10 −5 μm −3 , depending upon the specific particle detector used. This paper describes the operating principle of the AMC and the performance characteristics of a prototype device developed at the Particle Technology Laboratory, University of Minnesota.

Equilibrium bipolar charge distribution of aerosols

Abstract The equilibrium bipolar charge distribution of aerosols has been measured using monodisperse DOP (di-octyl phthalate) particles of 0.53, 0.55, and 1.17 μm diameter. The measured charge distributions are in good agreement with the Gaussian distributions predicted by Boltzmanns law. Additionally, the ratio of charged and uncharged particles was measured using monodisperse NaCl and methylene blue particles between 0.02 and 0.2 μm diameter. The result was compared with the theories of Bricard, Fuchs, and Natanson as well as with continuum and free molecular theories and values predicted by Boltzmanns law. Best agreement is again obtained for Boltzmanns law. The results indicate that the existing combination rate expressions for ions and particle are not sufficiently exact to account for the experimental results obtained here.