Issues with OpenModelica FMU export

Question

My model works as expected in OMEdit using DASSL, Euler and CVODE solvers. However after exporting it to a FMU, I get error messages and the simulation fails. I am using OMEdit v1.21.0 (64-bit).

My model is as follows:

model experiment 
  import Modelica.Units.Conversions.from_deg;
  import Modelica.Units.Conversions.to_deg;
  import Modelica.Constants.g_n;
  import Modelica.Constants.pi;
  import Modelica.Constants.small;
  parameter Modelica.Units.SI.Length Lw =10;
  parameter Modelica.Units.SI.Length B =3;
  parameter Modelica.Units.SI.Angle beta=0.35 ;
  parameter Modelica.Units.SI.Mass m =8000;
  parameter Modelica.Units.SI.Length xG =3;
  parameter Modelica.Units.SI.Length zG =1;
  parameter Modelica.Units.SI.Length fG =0;
  parameter Modelica.Units.SI.Angle eta =0;
  parameter Real U =18;
  parameter Real rho(unit = "kg/m3") = 1025;
  parameter Modelica.Units.SI.KinematicViscosity nu = 1.19 * 10 ^ (-6) ;
  parameter Modelica.Units.SI.Torque M = 0 ;
  parameter Modelica.Units.SI.Angle betaM = 0.01 ;    
 //Unknowns
  Real fnb ;
  Real clb ;
  Real cl0;
  Real lam(start = 0.9, fixed = false) ;
  Real cld ;
  Modelica.Units.SI.Velocity UM ;
  Real rn "Reynold's number";
  Modelica.Units.SI.Length lm ;
  Real cf ;
  Modelica.Units.SI.Area Af ;
  Modelica.Units.SI.Force Ff ;
  Modelica.Units.SI.Force rT ;
  Modelica.Units.SI.Force T ;
  Modelica.Units.SI.Force Fn ;
  Modelica.Units.SI.Length xP ;
  Modelica.Units.SI.Length aN ;
  Modelica.Units.SI.Length aF ;
  Modelica.Units.SI.Angle tau(start = 0.069, fixed = false) ;
equation
  fnb = U / (g_n * B) ^ 0.5;
  clb = m * g_n / (0.5 * rho * U ^ 2 * B ^ 2);
  clb = cl0 - 0.0065 * to_deg(beta) * cl0 ^ 0.60;
  cl0 = to_deg(tau) ^ 1.1 * (0.012 * lam ^ (1 / 2) + 0.0055 * lam ^ (5 / 2) / fnb ^ 2);
  cld = 0.012 * lam ^ (1 / 2) * to_deg(tau) ^ 1.1 - 0.0065 * to_deg(beta) * (0.012 * lam ^ (1 / 2) * to_deg(tau) ^ 1.1) ^ 0.6;
  UM = U * (1 - cld / (lam * cos(tau)) ^ (1 / 2));
  rn = lm * UM / nu;
  lm = B * lam;
  cf = 0.075 / (log10(rn) - 2) ^ 2;
  Af = B ^ 2 * lam / cos(beta);
  Ff = 0.5 * rho * UM ^ 2 * Af * (cf);
  M = T * fG - Fn * aN - Ff * aF ;
  0 = T * cos(eta) - m * g_n * sin(tau) - Ff;
  rT = (m * g_n * tan(tau) + Ff / cos(tau));
  0 = Fn * cos(eta) - m * g_n * cos(tau + eta) + Ff * sin(eta);
  xP = lm * (0.75 - 1 / (5.21 * fnb ^ 2 / lam ^ 2 + 2.39));
  aN = xG - xP;
  aF = zG - 1 / 4 * B * tan(beta);
end experiment;

I have simulated the FMU in both OMEdit and FMPY with the same error messages:

assert            | debug   | Model error: Argument of log10(rn) was 0 should be > 0
stdout            | warning | Non-Linear Solver try to handle a problem with a called assert.
assert            | debug   | Model error: Argument of log10(rn) was 0 should be > 0
assert            | debug   | Model error: Argument of log10(rn) was 0 should be > 0
assert            | debug   | Model error: Argument of log10(rn) was 0 should be > 0
assert            | debug   | Solving non-linear system 38 failed at time=0.
|                 | |       | For more information please use -lv LOG_NLS.

I am thinking whether this could have something to do with initialization?

Any ideas?

========================================

Edit 20.4.2023:

1. I tried (log10(rn+1000)-2) to make sure it is a positive value from the beginning but now get this error:

   assert | debug | hullResistance_12jac.c:274: Invalid root: (0)^(-0.5) assert | debug | hullResistance_12jac.c:274: Invalid root: (0)^(-0.5) assert | debug | hullResistance_12jac.c:274: Invalid root: (0)^(-0.5) assert | debug | Solving non-linear system 38 failed at time=0. | | | | For more information please use -lv LOG_NLS.

2. I did the FMU check as suggested and the model passes without errors.