English (unofficial) translations of posts at kexue.fm
Source

Scientific Space: An Interesting Type of Square Number

Translated by DeepSeek V4 Pro. Translations can be inaccurate, please refer to the original post for important stuff.

Numbers are beautiful and extremely charming, just as—

There is a type of number that, when split into two numbers, the square of the sum of these
two numbers equals the original number. For example: \begin{aligned} 2025&=(20+25)^2 \\ 88209&=(88+209)^2 \\ 152344237969&=(152344+237969)^2 \\ &... \end{aligned}

Below are some studies regarding this type of number:
  1. The essence of these numbers is: (A+B)^2=10^nA+B. For the general form (A+B)^2=kA+B, we have: A = \frac{k}{2} - B \pm \sqrt{\frac{k^2}{4} - (k-1)B} Therefore, generally, for a suitable B, two corresponding values of A can be found.

  2. For the case (A+B)^2=10^{2n}A+B, A and B can be: B = \frac{10^{2n}}{4}, \quad A = \frac{10^{2n}}{4} \pm \frac{10^n}{2}

  3. General Solution:

    For (A+B)^2=10^nA+B, we have (A+B)(A+B-1)=(10^n-1)A. Regarding this, we know that we need to find the product of two adjacent natural numbers that is a multiple of (10^n-1). Suppose we let 10^n-1=X \cdot Y and A=M \cdot N, and satisfy M \cdot X = N \cdot Y \pm 1.

    X and Y can be pre-determined from 10^n-1. By finding a suitable N, we can determine M. Thus:

    In the form M \cdot X = N \cdot Y + 1, we have A=MN and A+B=NY+1;

    In the form M \cdot X = N \cdot Y - 1, we have A=MN and A+B=NY.

    In this process, the Remainder Theorem and factorization will be used.

Examples:

3.1 Finding 2-digit A and B,
That is, (A+B)(A+B-1)=99A.
Let 99=9 \cdot 11 (It cannot be split into 3 \cdot 33 or 1 \cdot 99; think about why?). In the form M \cdot X = N \cdot Y - 1:

We have (11N-1) \pmod 9 = 0.

Thus N is a number of the form 5+9p. Taking N=5, we get M=6, A=30, and B=25. At the same time, for B=25, another A is 20. That is: \begin{aligned} (30+25)^2 &= 3025 \\ (20+25)^2 &= 2025 \end{aligned}

3.2 Finding 6-digit A and B,
That is, (A+B)(A+B-1)=(10^6-1)A.
Since 10^6-1=3^3 \cdot 7 \cdot 11 \cdot 13 \cdot 37, we can let 10^6-1=143 \cdot 6993 (not unique). In the form M \cdot X = N \cdot Y - 1:

We have (6993N-1) \pmod{143} = 0.

We can solve for N as a number of the form 51+143p. Taking N=51, then M=2494, A=127194, and B=229449. At the same time, for B=229449, another A is 413908. That is: \begin{aligned} (127194+229449)^2 &= 127194229449 \\ (413908+229449)^2 &= 413908229449 \end{aligned}

At this point, the problem is basically solved. Following this line of thought, we can find more such square numbers. Moreover, in general, the number of such square numbers is infinite.

Please include the address of this article when reposting: https://kexue.fm/archives/7

For more detailed reposting matters, please refer to: Scientific Space FAQ