A series of impact experiments were conducted on debris flow slurries and the mixtures of the debris flow slurry and large-sized particles with densities of 1 400-2 200 kg/m3

and 31 groups of impact experimental data at velocities of 2.4-5.2 m/s were obtained. The noise signals generated from device vibration and environmental interference are filtered out with the wavelet analysis method. Based on the results of the frequency spectral analysis using the fast Fourier transform method

a frequency threshold of 2 Hz is identified so that the low-frequency slurry impact force and the high-frequency large-sized particle impact force can be effectively separated. In the current hydrodynamic model the empirical coefficient α is generally difficult to determine. To resolve the issue

the functional dependence of the fluid Froude number Fr on coefficient α is developed based on 157 sets of debris flow monitoring data; and a formulation for calculating the slurry impact force is proposed

which represents different flow forms and the diminishing size effect. Compared to the smooth nature of slurry impact force

the large-sized particle impact force has more random characteristics. Both the number and frequency of particles impact on sensors increase when more particles are mixed in the flow. When mass ratio of particles increases from 0.05 to 0.21

the impact number increases from 1 305 to 2 838 times

and the impact frequency also increases from 82 to 195 times per second. The average particle impact force is about 60 kPa

which is about 3 times that of slurry. The frequency of particles detected by the upper sensors becomes larger than that detected by the bottom sensors as the particles content increases

demonstrating that the particles are prone to concentrate on the surface or at the head of debris.