Cisco Routing – How to Ping a Router with Same IP Address and Mask

Peter is correct.

Your router will look in it's ARP table for the MAC address associated with the IP address you've asked it to PING. Since the IP address is used by your 0/0 interface, there will always be a permanent ARP table entry for the IP address with the 0/0 interfaces' MAC address, so it will never even send an ARP broadcast, knowing the IP address is local.

If you can add two unique secondary IP addresses to both routers' interfaces, in a different (otherwise unused) subnet, you could ping the other router at that secondary address.

However, after answering your question as written, to expand on Peter's answer, if two devices in a subnet are using the same IP address, they will both reply to ARP requests, providing unpredictable results when other hosts try to reach the routers or anything beyond those routers. Based on timing of the ARP replies being received, packets may go to/through one router or the other. That may or may not be a problem (if any state-tracking devices are in the path, such as firewalls or VPN tunnels, it WILL be a problem), but it is definitely considered an incorrect configuration.

There is likely to be a better, more reliable way to do whatever is being attempted by using the same IP address on two routers - since you mentioned you're trying to make sure the other router works, you might want to look into HSRP (Cisco proprietary) or VRRP (industry standard), where the two routers have unique IP addresses (in your case, 72.40.40.2 and 72.40.40.3 would be good choices if available, or maybe 72.40.41.253 and 72.40.41.254, but any two unique IP addresses in the subnet), and share a "standby" IP address (72.40.40.1) that is used by clients connecting to/through the router(s). Only one router at a time uses the standby IP address, and they communicate to coordinate fail-over / fail-back.

If HSRP or VRRP is not available, or will not work, another method is to give each router it's own IP address and add both routers as equal cost routes on the devices that send traffic through them. If the routers are supposed to be the default gateway for client machines, this may not work, and it can cause duplicate packets to be sent out in some cases.

Going more esoteric, if the devices are Cisco switches that support stacking modules, and they're connecting to the same switch or another stacked switch, Multi-chassis Etherchannel might do what you're looking for redundancy-wise - it would be a way to bond both interfaces and give the port group a single IP address. VSS can do similar things in higher end switches.

Hard to know what direction to go in to resolve the multiple hosts with one IP address problem you've reported without knowing what other problem was being addressed to create this situation.