The aim of this paper is estimation of sensitivity of NA62 experiment towards the search for forbidden decay $\pi^{0}\rightarrow3\gamma$, which allows to test C invariance in electromagnetic interactions. Current upper limit of the probability of the decay is $3.1\times10^{-8}$ with $90\%$ confidence interval. $K^{+}\rightarrow\pi^{+}\pi^{0}$ decays are used as a source of $\pi^{0}$. The idea of selection of events with 3 products after $\pi^{0}$ decay in NA62 detector was developed. NA62 detector and all physical processes were modelled with Monte Carlo method using Geant4. Estimation of experiment sensitivity without consideration of background events is 1 order better, than current upper limit on the forbidden decay. Main background processes were studied. Main contribution in background events is $K^{+}\rightarrow\pi^{+}\pi^{0}, \pi^{0}\rightarrow2\gamma$ decay, when photon interacts with detector, and electron-positron pair is created. Specific veto was developed to suppress these background events. The signal region was selected. Comparison of Monte Carlo simulation and experimental data was done. The estimation of number of background events inside signal region was obtained. With strong selection criteria the lack of Monte Carlo events is observed. Analysis of background events is being continued in order to intensively generate more Monte Carlo events.
$^1$Institute for Nuclear Research of the Russian Academy of Sciences (INR RAS)