Maybe this will help you out...
http://www.apicsllc.com/apics/Misc/filter2.html
The link gives you a basic calculator, you just tell it the impedance of your drivers and what crossover frequency you want. You can do from 1st order ( -6 dB/oct) through 4th order (-24 dB/oct). It will tell you what inductor and capacitor values to use.
Now for the bad news... the results will be wrong. Simple calculators like this assume a flat frequency response from the drivers and they assume DC resistance from the speaker impedance. It may give you a starting point but if the results are good it will be due to shear luck.
The problem is that a speaker voice coil is not a resistor, it's a resistor in series with a capacitor in series with an inductor. Capacitors and inductors impedance vary with frequency. What's worse, the speaker inductor and capacitive elements are not constants. So, you have an element that varies with frequency even if it's value were constant and to make it worse it's value is not constant. They are variables within variables.
The other problem is that the speaker frequency response is not flat. Even if the impedance issue didn't exist you would still have the problem of such a simple crossover calculator not compensating for irregularity in the driver frequency response.
To do it right you need a measurement system to measure both the impedance curve (the impedance across frequency both magnitude and phase) and the frequency response. Then you need a real crossover software package like LspCAD http://www.ijdata.com/ (about $1,100.00) Once you have all that, you can do it right!
Baring having the proper equipment, being armed only with the simple calculator in the link above your only option is to do this by trial and error, mostly error. Start with what the calculator gives you and listen and adjust. If you download ARTA http://www.artalabs.hr/ and put a microphone with it you can at least get a look at frequency response. Then you can see what happens when you change values. With a simple inductor and capacitor for a 2nd order low pass filter on the woofer for example you might get a peak at the crossover frequency. ARTA would allow you to see this. In that case add some resistance (2 or 3 ohms to start) in series with the shunt leg (the capacitor in this case). Observe the peak go away.
In truth, with the ARTA measurement system and the simple calculator you will be better equipped than I was many years ago before going through many thousands of dollars in other measurement equipment far more expensive than LspCAD. It will be a bit of work trying different component values while watching the measurement result but in the end you can get good results.
You have your work cut out for you but it can be fun!
mk