The designer of the ACC should assure that fan pass frequency or fan blade natural frequency does not come within 10 to 20% of the motor bridge natural frequency. I believe the API suggests 20%.

There are two excellent presentations that discuss this topic on the ACCUG website.

The first is by N. Romano which was presented in 2015 no. 3 paper, Wind Loads on Fan Blades and Blade Dynamics. He discusses how not only do you think about the motor bridge but actually the prevailing wind side has downward flow and on the leeward side there is upwards flow. So as blades pass through this region there is a vibration inducement. For this reason, I favor off numbers.

The second presentation by Mr. Scottt Basham presented in 2014 as paper no. 10 named Maintenance and Vibration Issues and Solutions. They also experienced many vibration related trips an discuss how they solved this problem.

There is no rule for even or uneven number of blades and there are opinions favoring both options as some say an even number passes the bridge simultaneously and others say an uneven number cause a rocking motion.

Vibration trips indicate something much more intense than a fan passing the bridge IMHO. I would first inspect the blade pitch of the blades my notiong if all blades pass the same point on the fan casing when the fan is spinning very slowly.

Some fans have a tracking adjustment that bring an extra variable to the equation, just to be clear, despite the fact I am mentioning this does not mean that Howden fans require tracking adjustment.

If this does help, a vibrational analysis would seem in order to troubleshoot the issue.