Multiple Fact Table / Calculated Measure

Hi
I have the following structure currently a fact With Order details in and multiple dimensions (customer, date, product etc) in a star schema, I now have a requirement to show the cost of the job broken down by the sales person / employees percentage ownership associated with the order there are multiple types of ownership e.g. coordinator, sales person etc and there can be multiple people assigned to each order. I currently have the following design;

---------
---------
Order
---------
OrderPK,
Value
---------
---------

and am looking at the following new table

------------
------------
Order Owner
------------
------------
OrderFK
OwnerTypeFK
OwnerFK
OwnershipPercentage
-------------

and then creating a calculated measure in the cube which will work out the value * ownership percentage to show the value split against each sales person / employee however I am having problems when adding this data in to the cube as it is duplicating the values. I am using SSAS and have set the dimenions as many-to-many where referenced through the other fact table the only shared dimension is the Order dimension (and the employee dimension is referenced by both but this is not shared) can anyone offer any advise on where I may be going wrong? Is it possible to have calculated measues based on different measure groups? Or should I be looking at a different design completely?

It sounds like you're on the right track w/ the many-to-many design.  However, in the bridge table (Bridge_OrderOwner) used to link DimOwner to FactOrder, you'll want to include a measure (OwnershipPercentage) representing the percent-ownership and use it in a measure expression to deallocate the order value across the various owners.

Take a look at the Measure Expressions and Many-to-Many Dimensions section from this link.