What ever your workflow is if you submit a RGB file to QTR then R=G=B must be true otherwise those values are changed to calculate the K values since it is a gray scale RIP.
From the QTR user guide:
The .tiff format is very flexible and images stored in this format can have a
wide range of characteristics. Recent versions of QuadToneRIP have
removed many of the constraints on the image characteristics, but it must
be either a grayscale or RGB image. If the image is RGB, QuadToneRIP will
convert it to grayscale. If the image is 16 bit, it will be converted to 8 bit.
Alpha channels in the image will be ignored and layers are permitted. The
image can be in compressed (LZW only) or uncompressed format. Note that
any conversions are only applied to the image data actually processed by
QuadToneRIP.
When QTR does the conversion if a pixel value doesn't have R=G=B then QTR will convert those unequal values behind your back and the result may not be to your liking. So to repeat or to clarify, convert to Gray Gamma 2.2.
I would suggest you do some reading of the archives and other sources to understand the best practices for printing with QTR.